Larry Connors introduces his 2018 book, Buy the Fear, Sell the Greed: 7 Behavioral Quant Strategies for Traders, by stating in Chapter 1 that the book shows when, where and how: “…to trade directly against traders and investors who are having…feelings of going crazy and impending doom. …The goal of this book is to make you aware of when and why short-term market edges exist in stocks and in ETFs, and then give you the quantified strategies to trade them. …Thirty years ago, when a news event would occur, it could take days to assimilate it. …The only thing that’s changed is the timing of their emotion; today it occurs faster and at times is more extreme primarily due to the role the media (and especially social media) plays in disseminating the news that triggers this behavior.” Based on analyses of specific trading setups using data through 2017, he finds that:
From Chapter 2, “RSI PowerZones” (Pages 19, 21, 29): “RSI PowerZones are high-probability buying levels for ETFs, especially US Equity ETFs, when they are in an upward trend. …Four simple rules to trade SPY. …The strategy on average holds SPY from 3-7 trading days and buys when fear (and oftentimes a great deal of fear) is present. …Emotionally this is a hard (and even brutal) trade to take. The logical move is to ‘step aside and wait this one out.’ If you believe in buying the fear, though, you had the perfect scenario.”
From Chapter 3, “Crash” (Page 36): “Taking a CRASH trade is going to be very difficult for most traders because the stocks it shorts are usually moving parabolically. And many of these are being driven higher by speculators, the news media, social media, and massive short-covering. These stocks are very often ‘story stocks on steroids.’ And story stocks on steroids brings out the greediest worst in investors, traders, and speculators. …Sometimes these buyers are right and the stocks continue to go even higher. But more often they’re wrong…”
From Chapter 5, “Vol Panics” (Page 69): “We’re looking to ‘precisely measure’ the fear using structured rules that have historically told us that there is a great deal of fear in the marketplace, VXX is high, and historically shorting VXX at these measured levels has led to VXX prices declining over the short term the majority of the time. With the VXX Strategies, I’ll teach you two precise ways to do this. One holds a VXX short position for only a few days (it’s very short term). The other holds it longer, allowing both for the fear to subside and for the structural aspects of VXX to play itself out over an average of a few weeks.”
From Chapter 6, “VXX Trend Strategy”, (Page 84): “…the goal is to get short at the proper time, climb aboard as it declines, and stay aboard as long as possible. We also want to exit early if there’s a change in trend because if there is we can safely be in cash. How do we do this? It’s fairly simple: with moving average crossovers. …When the shorter period crosses below the longer period moving average, you go short VXX. When it crosses above its longer-term moving average you exit and go into cash.”
From Chapter 7, “Trading New Highs”, (Page 97): “We have a stock making a new 52-week high, oftentimes on good news (or anticipation of good news), and everyone feels good. …The 52-week high often attracts ‘breakout buyers’ who believe there is clear sailing ahead… At the same time, professional money (knowledgeable money) begins selling into the buying, having been smart enough to buy ahead of the good news and they now begin locking in their gains. In our rule-based scenario, we’re looking for the selling to then become extreme, eventually creating a form of panic selling.”
From Chapter 8, “TPS; Fear and Greed Rising“, (Pages 108-109): “TPS epitomizes the fear and greed aspect…especially with equity indexes around the world and especially with the main US equity ETF — SPY. …TPS identifies when an ETF is overbought or oversold and it then averages into the position as it becomes more overbought and more oversold. It then combines Time, Price, and Scale-In to enter a position as fear is rising, and on the short side, when greed is rising.”
From Chapter 9, “Terror Gaps”, (Page 126): “You want the security (and in this case we’re going to say ETF) to have sold off over the past few days. …You want the ETF to gap lower. Not only have the owners of the ETFs lost money over the past few days, they wake up and immediately get hit with a further loss. …You then want the ETF to sell off even further intraday. …Now the ETF is plunging further. Chaos is oftentimes running through their mind and they go into protection and survival mode (otherwise known as terror mode). At this point they’re most fearful and vulnerable and they are often simply selling irrationally. You now have the ability to buy an ETF from scared money.”
In summary, Buy the Fear, Sell the Greed offers traders a menu of tactical trading setups based on greed/fear hypotheses with quantitative backtests for each.
Cautions regarding findings include:
- As noted in the book’s disclaimer, results are gross, not net. Accounting for trading frictions would reduce reported returns.
- Strategies presented in the book are trade-level. Trade-level analysis differs from portfolio-level analysis. The former looks at events, while the latter tracks the value of a portfolio seeking to exploit a stream of events. Specifically:
- The portfolio must hold liquid reserves for exploiting a stream of events that is unpredictable in terms of both timing and overlap, so portfolio return is generally much lower than average event return. Events may be rare, such that the portfolio is mostly in cash. Employing multiple trade-level strategies exacerbates portfolio-level constraints. (For more, see Section 8.2 of “Chapter 8: Two Analysis Regimes”.)
- Average trade-level performance does not reveal portfolio-level drawdowns that may occur due to consecutive unfavorable trades.
- Portfolio-level constraints may help explain why some event-level anomalies persist.
- The tests presented are in-sample and involve multiple rules and parameter settings. Consistency of in-sample strategy performance over time is encouraging, but it is not sufficient to rule out luck due to survivorship bias, whereby a researcher winnows a large number of strategy variations by dropping those that fail a consistency test during the test period. An investor operating in real time during this period would not know in advance which ones to drop.
