key: cord-0060931-erlmtpif authors: Tata, Fidelio title: Research date: 2020-07-20 journal: Corporate and Investment Banking DOI: 10.1007/978-3-030-44341-2_6 sha: 23923602b0455c03e6a8488e02fdeb9c32ee7b36 doc_id: 60931 cord_uid: erlmtpif The market for global market research is outlined and the main providers of research discussed in detail. The chapter helps understand the various services provided by research analysts. A particular focus is set on trade idea, discussing the different types of trade ideas, general rules about them and process of generating them. Conflict of interests are addressed. The latest regulatory changes affecting research areas are discussed. The chapter concludes by listing the main challenges ahead. Following the financial crisis of 2007-2008, investment banks have been allocating significantly less capital to research as part of widespread cuts to restore profitability. According to some estimates, the number of analysts at the 12 biggest banks fell from more than 6600 in 2012 to fewer than 6000 in 2016. 2 Banks have also been cutting the cost per head by replacing experienced analysts with younger, cheaper staff, a trend labeled "juniorisation." Research analysts are under pressure to publish a large amount of research publications. 3 Most research is still distributed by email; email distribution groups are generously extended and often include recipients that have little interest in reading specific research publications. According to some estimates, more than 2 million different research publications are distributed every year, but only 2-5% of those are actually read on average. 4 Modern tracking methods allow research providers to monitor which research publications have been accessed by the clients, but banks are reluctant to disclose the precise percentage (most likely out of embarrassment). Investment research analysts can be categorized according to whom they work for. If they work for intermediaries that offer financial services to other institutional investors (brokers, dealers, investment banks, advisors, etc.), they are referred to as sell-side analysts; if they are employed by financial institutions that bear the economic exposure of the trades suggested by the research services, they are characterized as buy-side analysts; if they are offering research services independently and are neither part of the buy-side nor the sell-side, they are called independent research providers (IRPs). The tools used to conduct research are often similar, or even the same, for the three types of research analysts, yet the motivation differs. Sell-side analysts provide the buy-side with a wide range of research products, including thematic publications, trade ideas, market commentary, product and market research and technical analysis, hoping to increase deal flow with or gain advisory mandates from the client. Buy-side analysts conduct research to help the buy-side institution outperform the market and to create a distinguished investment proposition for the institutional investor's own clients. IRPs sell their research product to the buy-side for a fee. Sell-side research mainly serves one goal: increasing sales. Each business area within a sell-side firm instrumentalizes (monetizes) research for its own purpose. For example, a low interest rate view developed by research analysts in the Rates Strategy group will serve as a basis to suggest buying bonds in the bonds unit, receiving fixed on a swap in the derivative unit or buying stocks in the equity group. The performance of sell-side analysts is measured to a large degree against how it helps the sell-side firm to increase trade execution flow with clients. For example, if a sell-side strategist publishes a trade idea and a salesperson can confirm that this trade was executed with a client afterward, the analyst wins recognition. Also considered a positive contribution by Research is which a proposed trade idea does not get executed precisely in the terms laid out but can be used by Sales to initiate a client conversation that leads to other, related trades. This is harder to measure, but salespeople have a good sense about whether the day-by-day work of a particular research strategist helps them strengthening their client relationship and to increase business. As part of the annual performance measure process, research analysts are often asked to have a few salespeople provide feedback about the use of their research work. Sell-side research serves additional purposes, apart from increasing sales activity. The output of research units also has internal consumers, such as risk management groups, the Treasury department, accounting, proprietary trading desks, senior management and units responsible for regulatory reporting. For example, if the bank needs interest rate forecasts for their business planning process, Research is typically asked to provide the information. Sell-side research is also used by banks to showcase their expertise in products and markets. Research is posted (for free) on the bank's website, presented on conferences and client events, etc. Research then becomes a marketing tool. As a sell-side analyst, you may be assigned to different research areas depending on the bank's needs and your interests and abilities. To some degree, it is a process of trial-and-error. For example, at some point you will likely ask to accompany a salesperson on a client visit. Depending on how that goes you would be allowed to see more clients (salespeople share among each other experiences they had with research analysts). If you have good people skills and your presentations are appreciated by clients, you will gradually transition into a mostly client-facing position; otherwise, you will more likely be utilized in a more analytical capacity within the research production process. What is nice about the research environment is that it offers a home for people with almost any skill set. If you are very eloquent, big-picture (thematic) research could be your thing; if you a very technical, market micro-arbitrage trade generation would potentially be suitable. If you are a skillful writer, written publications could become your focus; if you more of a verbal person, presenting at morning meetings, during client calls or on conferences could be your strength. Sell-side areas can be differentiated according to various research disciplines. To just name a few dimensions: • Nearly all areas within capital markets and investment banking have their own "embedded" research unit, creating the risk of duplication, overlap, competition and inconsistency in research. A great deal of energy is spent within the sell-side institution on who "owns" the right to set the official prediction (house view, or "call") on a product, asset class, rate or price. Even when this was decided, some analysis may promote their own opinion as "risk scenarios" or "alternative views." A buy-side analyst is an analyst who supports the fund managers at mutual funds, pension funds, hedge funds, trusts and other buy-side institutions in the investment making process. The results of buy-side analysts are for the better part confidential and not for public consumption. The research conducted by buy-side analysts helps the buy-side firm increase the risk-adjusted return on their capital. Analysts are usually engaged in monitoring current news and trends, tracking down valuable information, compiling and evaluating research received from the sell-side, building proprietary financial models and conducting ad hoc requests by the buy-side traders. Because buy-side research typically does not get distributed outside the firm sponsoring it, buy-side analysts don't have to worry much about appearance, language and regulatory requirements (disclaimers, etc.) of what they produce; however, more rigor is applied to backtesting and risk analysis because buy-side research is often the basis for actual market transactions with an immediate P&L impact (while sell-side research is more often than not merely an invitation to consider a proposed idea). Because buy-side firms often hire seasoned sell-side analysts, starting your career as a sell-side analyst at a reputable sell-side firm may be advisable even if your goal is to ultimately become a buy-side analyst. Also, while working as a sell-side analyst you have the opportunity to network with a wide range of buy-side firms, allowing you to identify the best fit for you. The Centre for the Study of Financial Innovation (CSFI) defines IRPs as "standalone research firms, owned by neither buy-nor sell-side. Their clients are mainly asset managers and their USP is their independence, which avoids the conflict of interests that come with serving corporate clients. They receive most of their payments directly, in the form of subscriptions, payments for be-spoke research and consultancy, although some receive a sizeable amount via dealing commissions. The sector is fragmented -the research ranges from macro to micro and includes technology-driven modeling and analytics. Most are niche operators. The advantage is their focus and depth of knowledge; the disadvantage is the lack of diversified revenues and of a big balance sheet to tide them over in tough periods." 5 There are a number of reasons to expect that IRPs will be playing a bigger role going forward. First, the reputation of sell-side research has been harmed as conflict of interests have become more obvious to the public (see Sect. 6.7), especially after a group of investment banks agreed to pay roughly $1.4 billion in a settlement with regulators related to their behavior in rating technology stocks in the dot-com bull market of the 1990s. Second, as an increasing number of sell-side firms keep reducing the number of seasoned research professionals for cost reasons, experienced research staff is offering their specific expertise on an independent basis. Third, regulation has made the business model of selling research more attractive due to limitations imposed on using research as an inducement (Sect. 6.8). Whether or not investment research is providing much of a value is a hotly debated topic. On the one end of the spectrum are people suggesting that research (especially when coming from the sell-side) has only entertainment value 6 and does not help recipients to systematically outperform the market. The argument goes along the lines that markets are rather efficient to start with and even if there were temporary pockets of inefficiency, it would be highly unlikely that a research analyst identifies and publishes them before they have disappeared already. A somewhat cynical response of someone from this camp when receiving a research trade idea would be: "If you think this to be such a great trade, why don't you (or your trading desk) put it on yourself?". Of course, this point of view is rather extreme and ignores the fact that research is not just producing trade ideas but also helps compile and analyze data for decision makers on the sell-side, among other services. On the other end of the spectrum of opinions is the view that markets cannot become efficient without some market participants researching the market and taking advantage of dislocations, a notion related to efficiently inefficient markets discussed in Sect. 3.2.2. A balanced view between those opposing standpoints is maybe to assume that some research is useful, but not all. Like with other products and services, the principle of caveat emptor should be applied to investment research, meaning that because recipient of research assumes the economic risk it is imperative for them to always use their own judgment instead of blindly following research propositions. In the following section, we will discuss a number of research service offerings (mostly from a sell-side perspective) with respect to their value-adding, perceived or actual, to clients. Similar to what credit rating agencies (e.g., Moody's, S&P and Fitch) do for debt instruments, research analysts evaluate and rank equity instruments (e.g., stocks and corporate bonds). The evaluation typically includes some form of buy, sell or hold recommendation. This rating helps portfolio managers in their decision process regarding which securities of an otherwise diversified portfolio to over-and to underweight in an attempt to generate alpha. Research analysts' recommendations often carry weight with small investors as they tend to lack sophisticated financial data and seldom dig through corporate filings. Larger (institutional) investors have their own (buy-side) analysts performing ratings on existing and potential stocks in their portfolio; they often use sell-side ratings as a sanity-check for their own work. Upgrades and downgrades by research analysts often move stock prices because they imply that something has changed. They are often used to initiate a discussion with clients, often involving the research analyst(s) who worked on the research piece. As with many other research publications, the goal is often to give salespeople a good reason to call their clients. As will be discussed at greater lengths in Sect. 6.7, research analysts typically shy away from issuing sell recommendations because of conflict of interests. However, there is also a very practical reason why a buy recommendation can be monetized by Sales more efficiently than a sell recommendation: Sell recommendations can only be executed by clients that already own the security, while buy recommendation can be executed by anyone that is allowed and willing to trade in the security's asset class. 7 Often, salespeople call their clients on a daily basis (sometimes even multiple times a day); they need something to talk about! Being able to present an attractive trading opportunity is usually a good way to get a client's attention during which an experienced salesperson can extract some valuable information from the client (client's thinking process, type of trades conducted, existing positions, need for future transactions, etc.) that helps to strengthen the relationship and to generate some execution flow, even if completely unrelated to the trade idea used to initiate the conversation. Trading opportunities have to meet a number of qualifications to be suited to be shown to a particular client. The most important are: • Trade does not include instruments the client is not allowed to trade (e.g., long-dated bonds when talking to a money market fund manager); • Trade fits client's risk appetite and market views; • Trade idea is actionable (meaning that execution levels suggested in the research publication are in line with current market levels, even after bidask spreads and other transaction costs are taken into consideration, and that traders are willing to execute in the size suggested); • Trade idea is not completely meritless. Some investor types, like hedge funds and proprietary trading desks, are mandated with ferreting out attractive trade opportunities. Sell-side research analysts assist those investors by providing a constant flow of trade ideas. Some are merely observations of presumed temporary dislocations pointed out to suggest a possible convergence back to historical averages (so-called convergence trades), others involve a detailed, model-based analysis. There are some institutional clients that are reluctant to be called by salespeople unless a specific trade idea is presented. Trade ideas will be discussed in more detail in Sects. 6.4-6.6. Research analysts have, or are supposed to have, enough time to think outside the box, presenting big-picture ideas that help clients to enrich their thought process. Those publications center around a theme, such as inflation, global trade, political events, environmental issues. Often, institutional investors have to develop and present their own thematic analysis internally or to their own investors; thus, help from the sell-side to develop those ideas is almost always appreciated. Thematic research is mostly a reaction to the buy-side ignoring highly redundant research reports generated by almost every broker-dealer in a similar fashion, put out on regular production cycles even if there is nothing new to be reported (so-called maintenance research), or publications replicating what other already put out before ("me-too" publications). Essentially, the buy-side's request is simple: "Don't call me unless you have something interesting to say!" Thus, research groups now offer, or claim to offer, "new themes," "novel perspectives," "thought-provoking" arguments and "alternative views." If delivered as promised, it helps investment managers to position for shift in trends, structural changes in the economy, geopolitical events and other overriding topics, such as the impact of the COVID-19 pandemic (COVID-19). Some investors don't just want to read research about financial instruments, but also want to meet with the firms originating them. Specifically, institutional buyers of stocks and corporate bonds like to visit the corporation's headquarters or production facilities and speak to the corporation's management. Corporate customers of a bank, in turn, may have an interest to present themselves in a positive light to potential investors in their stocks and bonds. Research analysts are often entrusted with arranging this. Research analysts are arranging access of investors to corporate executives, so-called corporate access, in multiple ways: • Non-deal road shows: Analysts taking company executives to meet current and potential investors; • Investment conferences: CEO, CFO or other senior managers give presentations, followed by one-on-one meetings with investors; • Company visits: Analysts take investors to company's headquarters where they meet management; • Group meals: Analysts invite investors to wine & dine with company executives. Many investment banks track the number of times their research analysts take company executives on the road to meet clients and use the number to help decide analysts' annual compensation. Analysts who want corporate top executives to attend private events with their investor clients have to show they are brand ambassadors. However, it is difficult for an analyst to be considered a brand ambassador if they have a sell rating on a stock. This creates some serious conflict of interests for the research analyst. Starting some 20 years ago, broker-dealers have been increasingly giving clients access to web-based analytical research tools and databases, so that they can do their own research and analysis. The purpose of those tools is mainly to make sure that clients come back next time they have transaction needs, i.e., to make them sticky. Historically, access to those all-in-one platforms was limited to clients that reward the broker-dealer with a certain amount of deal flow. More recent regulation (see Sect. 6.8) limits most inducements of this kind. Analytical web platforms typically consist of several elements. First, they give access to an extensive database, consisting of reliable historical securities data from a wide range of asset classes, currencies and indices. Often, economic data are included also. Real-time or close-to-real-time data are provided for some time series as well. Second, analytical functions are offered to the user, allowing the user to perform calculations on price histories of individual securities or entire portfolios of securities. This could be as simple as calculating the spread (difference) between two security prices, or to show a moving average. Third, the platform will likely also give access to in-house models for risk management and relative value analysis. Those include mortgage prepayment models, derivative pricing models or term-structure model. Fourth, analytical platforms give the user the ability to display data in multiple ways. Elaborate graphing tools help identify market dislocations, track price relationships on an ongoing basis, create a graphical representation of the performance of a specific trade or monitor entire portfolios. Finally, web platforms typically also provide links to the provider's research reports, market commentary and video/ podcast content. Because some clients prefer to use their own graphical user interface (GUI) many banks' analytical research tool provide direct access to the platform's analytical components through an application programming interface (API). Also, an increasing number of platforms also provide trading capabilities (access to liquidity and trade execution). Those platforms collect the client's trading activity in one place, provide transaction oversight along with individual pre-trade, trade and post-trade activities. A transformative new trend impacting both, the buy-side and the sell-side, is the augmentation of investment decision processes by machine learning and artificial intelligence (AI) tools. Machine learning in the context of financial market analysis is a data-driven attempt to discover previously unknown relationships between prices and other market data with the goal of being able to forecast the market dynamics and to construct successful trading strategies. AI describes the ability of machines to exhibit human-like cognitive behavior. Many significant market participants are already devoting significant resources to those developments in computer science. For example, on the buy-side the investment manager BlackRock has established a lab for AI in Palo Alto, California. 8 On the sell-side, JPMorgan Chase's technology budget in 2019 is some $11.4 billion 9 and about a quarter of Goldman Sachs' workforce is made up by computer engineers. 10 Increasingly more important are data beyond what is typically used (corporate earnings, economic releases, etc.). Those data are referred to as alternative data and can be used for novel analysis such as: • Analyzing traffic through corporate websites; • Use of text analysis tools to analyze transcripts of earning calls; • Use of geolocation data from smartphones; • Analysis of credit card purchases and social media; • Satellite imagery machine learning and computer vision algorithms. Alternative data can be generated by individuals (e.g., billions of individuals constantly radiating data about what they are doing on social media), by business processes (e.g., credit card transactions) or by sensors (e.g., satellite data). 11 While alternative data may provide new insight, the length of alternative time series is often too short for identifying long-term trends that span over a whole business cycle. The inroads advanced technologies have made on research can be illustrated in the field of satellite imagery machine learning. Space tech has become much cheaper and venture capital has fueled commercial satellite launches. Computer vision techniques applied to satellite imagery allows for the identification of cars, buildings or changes in landscape. Several hundred remote sensing satellites are already launched into orbit. Data from space is on such a massive scale 12 that only AI can process this. With this new technology, it has become possible to predict retailers' profits by counting cars on the parking lots (see Fig. 6 .1) before the firm's annual/quarterly report is published, to forecast crops' yields, to estimate the world's oil inventory by detecting shadows on them, 13 to monitor activity and inventory of dry bulk cargo at ports or containers at ports, and to count ships to establish baseline activity and monitor for abnormalities. Research often plays an instrumental role in the onboarding process driven by Sales (see Sect. 4.2.1). Research capabilities are presented to prospect clients as a way to differentiate coverage from what the client already receives. Sometimes it is even a research analyst that initiates a client onboarding process. Clients tend to develop a strong relationship not only with their sales coverage, but also with some research analysts. If the research analyst moves to another broker-dealer, some of the clients covered by him/her may be willing to add the analyst's new employer to the list of broker-dealers execution is conducted with. Research can increase the market value of a bank. Some broker-dealers describe themselves as a "research house" and use research rankings as a unique selling proposition (USP). Research publications can be prominently displayed on a bank's website and research findings are often quoted in the press, creating free publicity that is often more effective than paid-for advertisement. Research can become an important part of the customer journey that typically begins with the awareness stage. As many broker-dealers compete for the attention of institutional customers, showcasing strong research capabilities can be a competitive advantage when trying to win execution business in an otherwise quite homogeneous service offering. Research contributes to client education (see Sect. 4.2.8) and seminars/ conferences (see Sect. 4.2.9) , organized by Sales. Research analysts tend to do most of the talking at those events (while Salespeople often focus on the social program following the formal presentations). Last but not least, research also provides valuable services within the organization to "internal clients," such as the Treasury or risk management departments. For example, banks have to run so-called stress tests (complying with IFRS 9 regulation 14 ) and may use different interest rate scenarios developed within the research group. Trade ideas are recommendations to consider a potential actionable transaction. Calling them ideas highlights that there are merely opinions and observations that are to be used to inspire recipients and to be the basis for a further discussion. However, the idea is presented such that it includes a market transaction that can be executed in the market. An analysis predicting that Japan's unemployment is going to fall is not a trade idea because there is no actionable trade; a specific Dollar-Yen transaction based on Japan's unemployment projection is. Trade ideas have various formats, are based on different type of information and serve different purposes. While it would be beyond the scope of this book to list them all, if that is even possible, a selection of most relevant types is discussed in this section. Most trade ideas come from the fixed income area. Carry and roll-down represent the expected return to hold a fixed-rate instrument position assuming an unchanged yield curve over a defined period. Carry represents the cash return determined by the spread between fixed payments of an instrument (e.g., the coupon payments of a bond) and the short-term funding cost (e.g., the overnight repo rate). Roll-down measures the expected price change of a position, assuming the fixed income instrument matures (rolls down) on an unchanged yield curve over a defined period. 15 Positive carry describes the situation in which a position gains in value if "nothing happens " (except the passage of time); negative carry indicates an expected loss in value in the absence of any changes. The empirical failure of the pure expectation hypothesis (stating that the effects of higher carry will be perfectly offset by a price decrease due to higher expected yields in the future, et vice versa, and that all bond positions have the same near-term expected return) appears to provide some justification for carry-and roll-down-strategies. Some investors have a strong preference for positions with positive carry. While they may have already established a number of trades that express a particular market view (requiring the market to move one way or another), they also want to hold trades that are expected to perform well if nothing happens. Those investors will then ask the research area for trade ideas that generate the maximum amount of (positive) carry and roll-down. Figure 6.2 shows an example of a three-month carry and roll-down calculation. Sometimes, carry and roll-down are risk-adjusted for comparison, due to the different volatility across the curve. Those trades are referred to as carry-to-risk trades and calculate carry-to-risk ratios to rank asset pairs across government bonds, swaps, credit, equities and currencies. Curve trades express a view about the steepness, or slope, of the yield curve. Since the steepness of a yield curve can be explained by many factors (expectations, risk premia, segregated markets, convexity, etc.), it is often not obvious whether a curve is too steep or too flat. Typically, curve trades are based on historical observations. "The curve has never inverted in this part of the curve " or "the last time the curve was this steep/flat was n years ago " are typical statements found in curve trade publications. Curve trades often bet on a reversal (convergence toward what is believed to be a "normal" or average slope). To avoid cash flows due to carry during the holding period of the trade, curve trades are often constructed in forward space. 16 If done, one needs to "beat" the forward slope (not the spot slope). The forward slope, however, is often less attractive (i.e., most of the expected convergence is consumed by negative carry). This means that there is actually no trade to be done, or the trade only works if one gets an almost immediate curve adjustment. Figure 6 .3 shows a comparison between a spot curve and a forward curve, representing market levels for US swaps observed in November 2006. Because the 1s10s 17 swap slope was much steeper in forward space, forward flattener trades appeared attractive. At the time, those trades were expected to perform well if the 1s10s slope would remain flatter than what forward slopes implied them to be in the future. The core principle of momentum trades is to identify, as early as possible, a pattern of market returns inconsistent with randomly moving prices. There is some empirical evidence supporting the notion of trending behavior in financial markets. If the presence of market momentum can be found, a trend-following strategy would suggest a trade constructed to benefits from a continuation of this momentum. One of the most common explanations for trending behavior is that markets do not respond instantaneously to new information as it takes time for information to get incorporate in market prices, leading to only gradual price adjustments. Momentum trades can be based on absolute or on relative momentum strategies. Absolute momentum strategies aim at identifying price trends of an individual asset, while relative momentum strategies rank assets on a relative basis (see Fig. 6 .4). Similar to momentum trades are trades based on seasonality. While momentum trades are based on trends in consecutive time periods, seasonality trades are based on assumed trends in non-consecutive time periods. An example would be a year-end trade based on the observation that the market is exhibiting a reoccurring price pattern at or around year-end. For example, many market participants are adjusting their trading positions prior to the year-end to show a more favorable balance sheet, to be published in the annual report. Such behavior could lead to price effects for certain Kolanovic and Wei [2015, 19]) assets. Apart from year-end seasonality, there could be quarter-end-effects, month-end-effects and many other possible seasonal effects. Fair value models are empirical models that attempt to quantify the linkage between macroeconomic fundamentals and market prices as well as between market prices themselves. They are not pure forecasting models, but require projections for economic and financial fundamentals to translate a view on the business cycle into fair values for assets consistent with this view. Fair value model trades are trade ideas based on a perceived deviation of market prices from model prices. Volatility trades are often based on a dislocation on the volatility matrix, which is a grid that summarizes implied volatility for various option maturities and tenures). Other factors being a basis for a volatility trade idea are changes in volatility skews (i.e., differences between at-the-money, in-the-money and out-of-the-money options) and dislocations between swaption volatility and cap/floor 18 volatility or between exchange-traded volatility and OTC volatility. Various investor classes transact in different volatility products, which can lead to relative value opportunities due to supply-demand imbalances; on the other hand, volatility products are typically priced off a unified model, which is constructed to be arbitrage-free. Thus, true arbitrage opportunities in volatility space are nowadays very rare. Some volatility products have a wide bid-ask spread (compared to, say, Treasury securities). Spotting volatility dislocations often help those investors required to buy volatility product (e.g., mortgage convexity hedgers) to identify cheap pockets of volatility. Technical analysis (TA) is a collective term for a number of quantitative methods to extract information from historical price patterns to forecast market trends. It includes chartism (pattern recognition), candlestick analysis, Elliott wave theory (dividing price paths into impulsive and corrective phases) and Dow theory (identifying sector rotations). The rationale of TA is hotly debated. Some people point out that it is a direct contraction to the efficient-market hypothesis, even in its weak-form version, because it assumes that there is information in historical data sets and graphical representations of them not fully reflected in the current market prices. Others point out that acknowledging some behavioral patterns in human investment decision making can help anticipate the future price evolution. Some market participants claim not to believe in TA per se, but are observing it anyway because they know that other people are believing in it and, by doing so, are creating a self-fulfilling prophesy. From a practitioner's perspective, it does not really matter whether TA works because it correctly predicts the future, or TA works because many people act upon the TA's prediction. For research, the discussion whether TA is a useful technique is a moot point. If clients want to receive it, it better be provided or else they will ask another broker-dealer for it. Not all clients want to receive TA, and some are even insulted by it, so it makes sense to produce it in stand-alone publications and to distribute it only to a selected group of investors. Some clients do not use TA as the sole and exclusive source when making an investment decision, but rather as one of many signals along with fundamental analysis, relative value arguments, flow-driven information, supply/ demand imbalances and other aspects. Some of what technical analysts are producing, such as simple pattern recognition in chart analysis, can be automated with the help of machine learning. It is fair to assume that in the long run the market won't need any longer an army of chartists to analyze the same data sets. Research analysts dedicated to observing, interpreting and predicting the US central bank are sometimes called "Fed watchers." Within the Federal Reserve (Fed), the Federal Open Market Committee (FOMC) is responsible for open market operations. It holds eight regularly scheduled meetings per year. At these meetings, the Committee reviews economic and financial conditions, determines the appropriate stance of monetary policy, and assesses the risks to its long-run goals of price stability and full employment. Changes to the Fed's target rate are most of the time announced on FOMC meetings (although they happen, on occasions, also between meetings). Predicting the Fed's next policy move leads to trade ideas. The best trade ideas are those expected to be profitable for a wide range of FOMC outcomes. There are central bank trades based on the expected behavior of other central banks as well, like trades based on the ECB or the Bank of England. Expressing a view on the central bank's target rate policy typically involves money market instruments (Eurodollar and Fed funds futures and futures options), while trades based on changes to Quantitative Easing (QE) involve longer-dated bonds, bond futures, bond options and swaps. Economic releases, such as the nonfarm payrolls release in the USA, can give important impulses for trading. Prior to the economic release strategist put out trades that are supposed to perform well if the data release matches their own expectation (or the official forecast/call of the broker-dealer). Often, investors have a view on data and request from research information about the following two questions: • What is the "reaction function" of various financial instruments to a surprise in economic data? • Which trades capture best any client-specific expected deviation from market consensus? Investment banks often employ one analyst dedicated to the tracking, interpretation and prediction of economic data releases. This analyst is typically co-located with sales and trading to comment on a data release right away (by shouting out loud on the trading floor or on the so-called squawk box or hoot-n-holler ("hoot" for short). In the minutes and hours following an economic release that significantly deviated from consensus, clients often experience adverse mark-to-market fluctuations on their trades and call the research strategist for an opinion about the further pricing dynamics. Conditional trades are trades that create an economic exposure if, and only if, a specified scenario materializes; they are constructed through option products. One class of conditional trades is conditional curve trades. They create a curve trade if interest rates move above or below a certain level. If, for example, a client faces curve risk, an offsetting curve position can be established conditionally; often, this is cheaper than an unconditional hedge. Another type of a conditional trade is a conditional swap spread trade. Swap spreads, being the difference between the swap rate and the government (Treasury) rate, are market directional. Often, investors are hesitant to express an outright swap spread widening view because of the risk of tighter spreads in a rally or in a roughly unchanged market. A significant rally is often accompanied by mortgage convexity hedging which leads to tighter spreads. A significant sell-off, on the other hand, often leads to wider swap spreads due to hedging flows and leveraged accounts expressing bearish views on the swap curve. So-called Put-Payers are conditional asset swap-widening trades in a sell-off. They are a combination of a bond futures put option) and a payer swaption. Typically, an investor sells a bond futures put and buys a payer swaption that matches optionality, cash-flow and strike levels. If the market does not sell off beyond the strike level (i.e., in a rally or an unchanged market environment), both options are expected to expire worthless without creating an economic exposure. If interest rates push above the strike level, both options are expected to be exercised. The investor gets delivered the cheapest-to-deliver (CDT) from the bond futures option, while simultaneously entering into a pay-fixed interest rate swap. Thus, the investor gets long the asset swap spread of the CTD at the strike-spread. Unfortunately, a large percentage of research "ideas" are just recycled recommendations that experienced institutional investors have already seen multiple times throughout their career. Sophisticated investors are not interested in wasting their time going through extended research publications in search for actionable suggestions-they want to see them right away to make a quick assessment whether they are worth further consideration. Trade ideas can play an important role as the focal point of an effective sales-client communication that saves clients time and helps "cutting to the chase." Vague recommendations are generally not helpful, as they lack specifics and often ignore negative aspects (negative carry, risks, illiquidity, etc.). Most investors expect precise details from a trade idea (and sometimes don't even want to hear any further explanation about the analyst's reasoning). While research recipients have different preferences with respect to the research format, there has been an evolution from essay-style toward trade idea-based publications (see Table 6 .1). Most institutional investors (with the notable exception of some hedge funds) face institutional restrictions as to what kind of trades they are allowed to establish. There are limitations with respect to the asset class (e.g., no derivatives), the maturity (e.g., only money market instruments up to one-year remaining maturity), the type of credit exposure (e.g., only investment grade), the envisioned holding period or the maximum risk (e.g., DV01 of a trade not exceeding a certain amount). To allow investors to quickly check those restrictions, trade ideas need to highlight the relevant specifics prominently. 19 Table 6 .2 lists the main elements that need to be part of a trade idea. Price level at which trade should be unwound to limit (cut) the loss The ability to come up with good trade ideas, possibly even on a daily basis, is a great value to any research analyst. Some are actually keeping their cards close to their chest and are reluctant to share their knowledge, out of fear they could lose their distinctive features within the pool of research analysts. 20 Also, the market is quite efficient in the sense that any framework developed to spot profitable trade opportunities will only work for a limited period, as more and more market participants start replicating the process and price inefficiencies are "arbitraged out." Thus, as a research analyst you have to constantly look out for new ways to ferret out interesting trade ideas. Conceptually, there are four generic methods to generate trade ideas. The first is a data-driven approach. Here, large sets of data (closing data pulled from the bank's databases or, even better, intra-day and tic-by-tic data) are analyses with respect to any possible dislocation. Dislocation may be defined as a, say, 2-standard deviation move from historical averages. The analytical framework is likely set up to create near-risk-free positions that tend to remain relatively stable and are widely independent on what the overall market is doing, something said to have little market directionality. An example for such a relative value position is a position on one part of the yield curve in comparison with positions in neighboring points on the yield curve. Depending on the resources available to the analyst, the analysis may be run on an Excel spread or as dedicated programs developed and maintained by IT departments. Once the analyst is alerted by the model that there is a potential dislocation (and after checking that is wasn't just due to data issues), the second step would then be to understand what has caused the dislocation, why it is assumed to be temporary and what will cause a convergence back to historical relationships. The second strategy in finding trade ideas turns the above process upside down. Since dislocations are typically caused by an imbalance in flow, one would try to identify one-way trading flows initiated by significant market participants. Research analysts following this strategy would literally walk down the aisles of the trading floor and question salespeople whether they have noticed unusually large flows. Once those flows have been identified, one would then check whether they resulted in a price dislocation. The advantage of this approach is that one does not have to scan through the entire universe of possible relative value relationships and can focus on relevant areas. Also, knowing what has caused the dislocation, if there turns out to be one, makes it much easier to determine whether this is a short-dated effect or a long-term structural change. The third way to generate trade ideas is to monitor what type of trade ideas other research analysts are publishing, 21 to check whether they are still attractive given current market levels and, if so, creating a variation of it that is then promoted under one's own name. The majority of trade ideas are created that way, which also explains why the buy-side is bombarded with similarly looking trade ideas from different sell-side firms every day. The fourth strategy is to reverse-engineer the thinking process of market participants that do not advertise their strategies but are known for executing profitable trades. Those are mostly hedge funds, proprietary trading desks and other relative value traders. Salespeople are often too busy executing orders to extract the informational value of the flow from those clients, but a knowledgeable research analyst with enough time at hand can infer from the different legs of a structured trade the underlying trade idea. However, arbitrageurs are well aware of this and frequently make detection of their trading strategy more difficult by executing the different legs of the relative value trade with different broker-dealers. Once a trade idea has been generated, it is necessary to prepare for questions from the investor side. Questions may include: • "How come no one else thought of this idea before?" • "Why aren't other people putting on this trade?" • "What has caused the dislocation in the first place?" One should also be able to explain what caused the dislocation and why this force is about to fade away. Reasons for a potential dislocation are: • Market is wrong and you are the first one having found about it; • Others know about the dislocation but cannot take advantage of it (e.g., for legal reasons); • Other market participants are causing the dislocation by trading the other way (for a good reason); • You are wrong (i.e., there is no dislocation). It is typically very convincing if one can demonstrate that one's own prop desk/trading desk has put on the trade already: "Our own desk has put on this trade in size already, but we maxed out on risk and want to give our favorite clients a chance to participate in the trade." Investors are suspicious individuals (as they should be). Many have had their share of bad experiences and only trust research analysts (and salespeople & traders) they have a long-term relationship with. To the question "Why are you showing me this trade idea? " they don't want to hear any of the answers listed in Table 6 .3. 22 We already discussed the general concept of principal-agent problems causing conflict of interests from a theoretical point of view in Sect. 3.2.4. More than just being an academic matter, conflict of interests in Sales have been very obvious to most market participant all along. As mentioned in Sect. 6.3.1, the observation that only so few sell recommendations are issued by research analysts is casting suspicion. Assuming that stock prices follow, at least approximately, a random process, buy-and sell-ratings should be issued in roughly equal frequency. Studies show, however, that merely 6% of research analysts' recommendations on stocks in the S&P 500 index are sell or equivalent ratings. 23 The fact that research analysts only rarely issue sell recommendations and forecasts are often excessively optimistic has led to questions about the robustness and quality of investment research. 24 It also led to regulatory Answers to "Why are you showing me this trade idea?" nobody likes to hear: "I am a junior analyst and nobody else is interested in listening to my recommendations yet. You are the beta-tester of my idea" "Smart money is doing exactly the opposite, but we need some dumb investors providing liquidity for them to get into the trade" "Smart money already put this trade on a while ago, made good money, and want to get out of the trade now. We are looking for some uninformed investors providing liquidity for them to get out" "Our trading desk is axed to do this" enforcement action. For example, in 2016 the Securities and Exchange Commission (SEC) charged a former Deutsche Bank research analyst 25 with certifying a rating on a stock that was inconsistent with his personal view. 26 The SEC investigation established that although the analyst recommended selling stocks of Big Lots in conversations with several hedge fund clients he didn't downgrade Big Lots from a buy recommendation in his report because he wanted to maintain his relationship with Big Lots management. During an internal conference call, the research analyst justified his behavior by saying "we just had them in town so it's not kosher to downgrade on the heels of something like that." Most large sell-side firms, besides producing investment research, also feature an Investment Banking Division (IBD). This area performs primarily advisory functions and includes Mergers and Acquisitions (M&A), Equity Capital Markets (ECM) and Debt Capital Markets (DCM). In order to increase the chance of winning (profitable) investment banking mandates it is tempting for the bank to win the clients favor by portraying the client in a positive light in research publications. Even after a corporate finance mandate has been secured, there is a benefit from "talking up" the value of a mandate's business as it increases the chance of a positive outcome (merger, acquisition, Initial Public Offering [IPO], etc.), increases the business deal's fees by inflating its market value and generates buying interest in the secondary market, which facilitate the profitable unloading of the banks own positions in the security. There are two ways a sell-side firm can address conflict of interests within Research. The first is to create a so-called Chinese Wall between Banking and Research. The term Chinese Wall (or just "Wall " for short) is inspired by the Great Wall of China and describes the (previously only) ethical (and now also) legal obligation to create a boundary between insider information gathered during an investment banking process and information used in investment research. There are strict regulatory rules governing the kind of information a research analyst is allowed to receive. The second way to mitigate conflict of interests is to isolate research strategists within a stand-alone research entity. An example for this is Kepler Cheuvreux, a leading independent European financial services company specialized in advisory services and intermediation to the investment management industry. Research analysts and research management are owning 31.9% of the equity. Among the minority shareholders are the sell-side firms Crédit Agricole CIB, UniCredit, Rabobank and Swedbank (see Fig. 6 .5). If you ever assume the role of a research analyst and keep defending your own integrity, along with the integrity of the institution you are working for, you will develop a reputation in the market that will benefit your career in the long run. It is primarily your responsibility to become aware of potential conflict of interests and to escalate them with management if they occur. Often, responsibilities can be realigned in quick and simple way that realigns your interest with that of the bank and its clients. The Markets in Financial Instruments Directive of 2014, 27 commonly known as MiFID II, aims to ban any inducement for fund managers to act in a way that is not in clients' interests. 28 MiFID II for investment research states that brokerages providing both research and execution services will need to supply and price them separately. Any benefit received by the investment manager must enhance the quality of service to the client (asset owner) and does not distort or bias the provision of services. One aim is to combat the perception that research is provided by investment banks and brokers to asset managers as an inducement, or bribe, to trade with them. Research in the context of MiFID II can be roughly defined as investment material or services concerning one or several financial instruments or other assets, the issuer or potential issuer of financial instruments, or something closely related to a specific industry or financial market. More specifically, Certain type of research provided to certain investor types needs to be priced to comply with MiFID II regulation. But what is the right price? The problem is that investment research had been given away for free in the past, so there is little knowledge of its price elasticity of demand. 29 Two possibilities are illustrated in Fig. 6.6 . Increasing the price of research from zero could cause a gradual reduction in demand, or demand could collapse if even only a small price is charged. Both academics and practitioners have been struggling, and still do, to establish a clear understanding of what the involuntary unbundling of research and execution services does to the research market dynamics. When looking for other markets where similar unbundling took place, some evidence can be drawn from the market of plastic bags (that used to be distributed freely by retailers as part of a service bundle). Aiming to reduce the waste of resources, laws in several countries forced retailers to stop offering plastic bags at a price of zero and to price them separately. Price elasticity of demand was found to be very high and demand dropped. 30 Is investment research the new plastic bag? Research used to be the first career stop for many new hires in Global Markets. Spending a year or two in research was considered to be a great way to acquire a broad knowledge about financial markets, products, clients and various coverage areas. Afterward, many research analysts would go on to work in sales, trading, risk management or other areas. It used to be quite easy to identify sympathetic employees from non-research areas and to have them help out in a research-related task. "I used to be in your shoes" was something along the lines they would say. Research no longer plays the same role of super-charging new hires due to shrinking research departments and for cost reasons. As a research analyst, it is now more difficult to find former research employees within the organization. An increasing number of research publications (especially those considered to be maintenance research) are getting outsourced and/or automated. Fewer seasoned research staff is needed for this type of research and experienced research analysts are replaced by junior, unexperienced employees. Research analysts now must either have the necessary data science talent to facilitate automation (e.g., knowledge of the programming languages Python and R, skills in cloud computing 31 ) or have strong people skills to deliver customized research in a client-facing environment. Regulation is making the research area more challenging also, as discussed in the previous section. Before MiFID II came into effect, it was not atypical for many sell-side research analysts to spend a considerable amount of time and energy on self-promotion, trying to create the impression that their research is essential for their employer to maintain profitable client relationships. Since 2018, clients have started voting with their wallet about which research they actually like to receive. This is nothing short of a total game changer in the area of investment research. Although this is a long process to play out, 32 eventually the degree of value-adding of specific investment research services will be less of a matter of opinion and something that can be measured. Chaigneau) The quarter century that I have spent in investment research, on both the sell and the buy-side, has seen extraordinary changes in the way research operates, although the core objective of the job hasn't really: It is still about delivering good investment ideas. Already at the time my own career started, Alan Greenspan, who served for nearly twenty years as the Chairman of the US Federal Reserve, was famous for his obsession with data: He was untiringly searching for on-the-ground evidence that would give him a faithful representation of the economic conditions. Since then data has exploded; so much that privacy has become a major challenge for business and society. Superior data provides an obvious edge to whomever has access to it and manages to use it effectively: that is true for policy makers as well as investment professionals. There is so much data now-free or not-that it is increasingly difficult for human beings to process it. That task is better left to the machine, hence the unstoppable rise of AI. Hence, Research has evolved in a way that now gives a much bigger role to quantitative analysts, data scientists and programmers. One typical transformation in the making is the shift from VBA-based Excel-spreadsheets toward coding in the programming language Python. This evolution matches that of the industry, characterized by the rise of automated and algorithmic trading, fintech and passive asset management. The latter development is profoundly transforming the buy-side, and the self-sustained trend very much relates to the role of research. If active portfolio managers-often supported by the work of traditional analysts-consistently fail to outperform market indices, it is hard for them to justify being paid large fees. Both the regulators and market forces-clients looking for a better returns and lower fees-contribute to the move toward passive investment, which is much cheaper to run. The falling fees lead to cost cutting, and a broader move toward automation. Pursuing a career in Research opens many doors, as there is a direct links with other positions such as a developer, trader or portfolio manager. The beauty of starting in Research, event for a quant, data scientist and programmer, is that you may be asked to collaborate with other analysts with a very diverse set of skills. From them you may learn about how capital markets actually work (the products and structures that are traded, the actors, the drivers, etc.). Eventually, a well-rounded research analyst with a decent understanding of capital markets and a strong expertise in quantitative analysis and/or technology has likely a bright future in the industry. For all that is said about the rise of passive investments and the failure of active asset management, there is still very much a role for traditional (some will say old-fashioned) analysts. In fact, the diversity of Research positions is remarkable, as the author of this book has shown. One may categorize them in three buckets: quantitative, top-down and bottom-up analysts. I will not say more about the quantitative analysts, the rising stars of an industry, irremediably caught in a race to the bottom (falling margins and fees). Top-down research, or global macro, includes economics, strategy and asset allocation. Bottom-up refers to equity or credit analysis, that is focused primarily on the business strategy and balance sheet of corporations. This is still at the very heart of active portfolio management. My conviction is that asset management is being "barbellized." Passive portfolio management will be dominated by firms that manage a huge quantity of assets (the BlackRocks and Vanguards of this world) on which they charge very small fees. Active portfolio management will be dominated by much smaller but agile boutiques that are specialized in specific strategies; the much smaller size allows them to deploy trades that the giants are simply too big to execute in areas where market depth is insufficient. Those stuck in the middle of that barbell will be squeezed. The traditional analysts of course are also challenged by technology. Credit and equity analysts for instance face the competition of AI-tech companies offering products that generate signals about business or rating trends are burgeoning. Those AI signals may emanate from social media feeds, data from financial reports, text analysis, etc. Those however are likely to complement rather than replace the role of analysts who talk with the management of those companies, understand the industry trends and interact with investors. Take an insurance company for instance: With regulation having forced a reduction of duration gap, the quality of the balance sheet to a large extent depends on the credit where it is invested; those that can rely on a solid credit research will tend to better absorb shocks in the cycle. Another example of competing or complementary developments lies in equity strategy. Traditionally, the strategists were focused on geographical and sectorial relative value and allocations. Nowadays, the equity market is increasingly dominated by factor analysis (value, growth, momentum, quality, etc.), which essentially is a quantitative activity. Again, there is still a role to be played for the traditional top-down analyst, and even more so if he or she is able to complement the own work by that of quantitative strategists. Finally, it is impossible to write about the exciting world of research without mentioning environmental, social and governance (ESG), criteria used by socially conscious investors when screening investment opportunities. The whole financial industry is adjusting to what has become both a purpose and a requirement (the ESG brand is attracting increasing inflows). This area is still developing, given the huge amount of data available and the flurry of providers, none of which has established itself as the new industry standard. ESG is also an area where fundamental analysis will assist the machine, or vice versa. While large segments of finance are being commoditized, let me finish with two important trends. First, there is the rise of private capital markets. The decline in interest rates has led investors to look for new investment opportunities, such as private equity and private debt, where they can capture an additional illiquidity premium. Those markets, being far less transparent and liquid compared to public markets, require different skills. Second, while electronic trading is dominated by the machines, many long-term investors still very much rely on thematic research. Secular stagnation, the saving glut, globalization (and now reshoring?), inflation (or the lack thereof ), unconventional monetary policy are all examples of themes that have truly shaped Global Markets for the past quarter century. Research remains an area of great diversity, where very different individuals can flourish, for as long as they have a passion for finance. be priced in the current yield environment. The expected roll-down is sometimes estimated by taking the difference between the current spot rate and a shortened rate on the same spot curve, and then multiplying this by the forward duration of the instrument. 16. "In forward space" means establishing a forward position. For example, instead of establishing a 2-year-vs.-10-year curve slope position using a 2-year and a 10-year spot-starting swap, it is constructed with a 3-month-into-2-year and a 3-month-into-10-year swap. This would then been called "putting on a 2s10s curve trade, 3 months forward." 17. 1s10s is short for 1-year-vs.-10-year, the yield difference between the 10-year rate and the 1-year rate. 18. An interest rate cap (floor) is a derivative instrument in which the holder receives periodic payments if interest rates exceed (fall below) an agreed-upon level. 19. Also, some investors expect the sale force to pre-scan research notes and to only forward those that do not violate restrictions or are inconsistent with their investment style. 20. There is even a story about a strategist who used to keep his relative value trade-detection spreadsheets on a USP stick that he would take home at the end of each workday, suspecting that if he were to save them on his work PC someone else would "steal" them from him. 21. This requires having access to competitors' research publications, which is getting increasingly more difficult. However, clients and prop desks are sometimes willing to forward research they receive from "the Street." 22. A research analyst with a strong relationship to a client may, however, playfully and jokingly suggest such an answer to tease the client, only to offer a convincing alternative explanation afterward. 23. Ng and Gryta (2017) . 24. See, for example, Galanti and Vaubourg (2017) . 25. The analyst agreed to settle the charges by paying a $100,000 penalty and was suspended from the securities industry for a year. Deutsche Bank agreed to pay a $9.5 million penalty to settle civil charges. 26. Source: U.S. Securities and Exchange Commission (2016). 27. European Union (2014). 28. MiFID II came into force in 2018. 29. Price elasticity of demand, or PED, is a microeconomic concept that measures the percentage change in the quantity demanded of a good resulting from a 1 percent increase in the price of that good. 31. Parallel, distributed computing on shared, remote-access computing and storage resources. 32. Many sell-side firms are only offering research bundles, so it is currently not always possible to tell what the demand for specific publications or for publications of one specific research analyst is. CRISIL Coalition This includes daily, weekly, monthly, quarterly and annual reports, stand-alone/thematic research, trade alerts and regional/sector compilations Financial market research can be quite repetitive, causing many research analysts, including myself, to seek out creative, funny or even poetical ways to transport otherwise uninspiring information Assuming clients are not engaged in so-called short selling, which is true for the majority of buy-side firms. Short selling is the process of borrowing a security from another market participant, selling the security in the market and then having to buy it back later to return the borrowed security to its lender Even a relatively small $200,000-satellite is capable of sending several terabytes of data on a daily basis Most of the fuel tanks have floating top so it is possible to use trigonometry to calculate its volume and the amount of oil in it IFRS 9 refers to an International Financial Reporting Standard that replaces the previously used International Accounting Standard (IAS) 39. Since 2018, regulatory-mandated stress testing needs to be conducted following the IFRS For example, the 1-year roll-down of a 5% 10-year bond currently trading at par would calculate how much above/below par a 5% 9-year bond would References Convery, Frank, Simon McDonnell, and Susana Ferreira Dimon Sounds a Cautious Note as JPMorgan Prepares for Recession. Bloomberg Directive 2014/65/EU of the European Parliament and of the Council of 15 May 2014 on Markets in Financial Instruments and Amending Directive 2002/92/EC and Directive 2011/61/EU Text with EEA Relevance. Official Journal of the European Union, L 173/349 Computer Engineers Now Make up a Quarter of Goldman Sachs' Workforce. CNBC A Level Playing Field for Investment Research? Challenges Facing the Buy-Side, Sell-Side and Independents. Centre for the Study of Financial Innovation publication Optimism Bias in Financial Analysts' Earnings Forecasts: Do Commissions Sharing Agreements Reduce Conflicts of Interest? Document de Recherche du Laboratoire d'Économie d'Orléans Big Data and AI Strategies: Machine Learning and Alternative Data Approach to Investing. JPMorgan Global Quantitative & Derivatives Strategy Momentum Strategies Across Asset Classes: Risk Factor Approach to Trend Following. JPMorgan Global Quantitative & Derivatives Strategy New Wall Street Conflict: Analysts Say 'Buy' to Win Special Access for Their Clients Some Poetical Thoughts on Bank Liquidity Price Formation of FICC Research Following MiFID II Unbundling Rules SEC: Deutsche Bank Analyst Issued Stock Rating Inconsistent with Personal View Final Call for the Research Analyst? Financial Times BlackRock Bets on Algorithms to Beat the Fund Managers Bag Elasticity. ThinkProgress