Hedge Fund Market Wizards
by Jack D. Schwager
Highlights
Dedication
“There is no adrenaline rush. If I get a rush, it means that something has gone horribly wrong. . . . The whole thing should be pretty slow and controlled.—Alex Honnold” p9
Preface
So much of what passes for high returns merely reflects a willingness to take more risk rather than being an indication of skill, I believe that return/risk is a far more meaningful measure than return alone. p13
It is ironic that in terms of any type of risk measure, hedge funds, which are widely viewed as highly speculative, are actually much more conservative than traditional investments, such as mutual funds. It is primarily as a consequence of lower risk that hedge funds tend to exhibit much better return/risk performance than mutual funds or equity indexes. p14
“Traders who are successful over the long run adapt. If they do use rules, and you meet them 10 years later, they will have broken those rules. Why? Because the world changed.”—Colm O’Shea p15
Part One: Macro Men
Chapter 1: Colm O’Shea
O’Shea views his trading ideas as hypotheses. A market move counter to the expected direction is proof that his hypothesis for that trade is wrong, and O’Shea then has no reluctance in liquidating the position. p19
To use a sailing analogy, the wind matters, but the tide matters, too. If you don’t know what the tide is, and you plan everything just based on the wind, you are going to end up crashing into the rocks. That is how I see fundamentals and technicals. You need to pay attention to both to make sense of the picture. p24
When I went to university, economics was taught more like philosophy than engineering. Since then, economics has become all about mathematical rigor and modeling. The thing about mathematical modeling is that in order to make problems tractable, you need to make assumptions. Assumptions then become axiomatic for the entire subject—not because they are true, but because they are necessary to get a solution. So, it is easier to assume efficient markets because without that assumption, you can’t do the math. The problem is that markets aren’t efficient, but that fact is just conveniently ignored. p24
The mathematization of economics has been a disaster because it has greatly narrowed the scope of the field. p25
I think the biggest mistake people make is to assume there is an answer when, in fact, there may not be a good answer. p26
All solutions that will work in the real world have to embrace the fact that the U.S. is not as rich as Americans think it is. Most political solutions will be in denial of that fact. p26
All that Soros did was to recognize that the situation was untenable. The Bank of England’s effort to support the pound was the equivalent of trying to fight gravity. p27
You can wait until people start to care. Taking Nasdaq as an example, you want to be selling Nasdaq at 4,000, but only after it has gone to 5,000. So you are selling the market on the way down, not on the way up. Because in a bubble, who is to say how far a market can go. p28
What is the one thing you know about a company that posts smooth earnings every single quarter? They are manipulating the numbers. Yes. You know nothing beyond that. They may or may not have a good business, but you know they are manipulating numbers. People love stable earnings. p28
We recognized that we would underperform the bulls by quite a bit because in a bubble the true believers will always win. That’s fine. You just need to make decent returns and wait until the market turns. p29
What I believe in is compounding and not losing money. p29
We did a lot of our trades through options—positions like buying calls in currencies with a carry because the positive carry paid for your option. Carry currencies are currencies with higher interest rates. p29
By being long options, you can never have a major drawdown. If the bubble continues, you make nice returns; if it collapses, you just lose the premium. p30
I try to avoid conceptualizing the market in anthropomorphic terms. Markets don’t think. Just like mobs don’t think. p31
Hedge funds are tiny in the macro space. If you are talking about tech stocks, then sure, hedge funds are massive. But if you are talking about the foreign exchange (FX) market or Treasuries, hedge funds are tiny compared to real money. In comparison, PIMCO or the Chinese are enormous. There are trillions of dollars moving in these markets, which make the little people like me quite irrelevant. We are not a force in pricing. One reason I like macro so much is because I am a small fish swimming in a sea of real money. Fundamentals matter. I am not playing a game against people like me. That would be a zero-sum, difficult game. p31
I didn’t need to know why. Once you realize something is happening, you can trade accordingly. Trades don’t have to start based on fundamentals. If you wait until you can find out the reason for the price move, it can be too late. A great Soros quote is “Invest first; investigate later.” You don’t want to get fixated on always needing a nice story for the trade. I am an empiricist at heart. The unfolding reality trumps everything. p32
Fundamentals are not about forecasting the weather for tomorrow, but rather noticing that it is raining today. The great trades don’t require predictions. p35
Equity markets would eventually notice, but being short equities is a hard trade because they might still keep going up for a long time. After a bull market that goes on for years, who is managing most of the money? The bears are all unemployed; they’re not managing any money at all. p35
The managers who are relentlessly bullish and who buy more every time the market goes down will be the ones who end up managing most of the money. So, you shouldn’t expect a big bull market to end in any rational fashion. p35
There will always be something that happens at the same time. Calling it a catalyst isn’t very helpful in explaining anything. Did World War I start because the Archduke was assassinated? Well, kind of, but mainly not. I don’t subscribe to the catalyst theory of history. But most people love it, especially in markets, because they can point to that one cause and say, “Who knew that could happen?” p36
If you play roulette, you are in the world of risk. If you are dealing with possible economic events, you are in the world of uncertainty. If you don’t know the odds, putting a number on something makes no sense. What are the odds of Germany leaving the euro in the next five years? There is no way of assigning a probability. p37
When you say you bought volatility, does that imply that you were not playing for any directional move in currencies? Yes, our assumption was only that it would move somewhere. So you put on positions like long straddles and strangles in currencies? Yes. p37
Corporate bonds pay a higher interest rate than U.S. Treasury notes to compensate investors for the higher risk. The yield difference is called the credit spread. The lower the individual bond rating, the wider the spread. Overall credit spread levels will widen during times of financial crisis when investors will demand higher interest rate differentials for accepting the greater risk implicit in holding corporate bonds instead of Treasuries. p38
Who is the natural seller of equity puts? No one. Who is the natural buyer of equity puts? Everyone. The world is long equities, and people like owning insurance, so there is an excess of natural buyers for equity puts. That is why equity option prices are structurally expensive. p40
You need to implement a trade in a way that limits your losses when you are wrong, and you also need to be able to recognize when a trade is wrong. George Soros has the least regret of anyone I have ever met. p44
You need a method that suits your personality. I don’t believe that I am an amazing economist who predicts the future. What I actually believe is that I recognize the world as I find it and that I am flexible enough to change my mind. p44
Why aren’t you a fan of trading equities? Interest rate markets or FX are usually better ways for me to express trades. The world is full of people who trade equities. I don’t think another hedge fund that trades equities is particularly exciting. Also, a problem I have with equities is that equity stories make no sense to me. Equity people often make no sense to me. The reasons I think trades have worked are usually nothing like the reasons why equity people think they worked. In my entire life, I’ve personally only done one single-stock trade. p46
Most trading book rules are designed for people who have the error of excess optimism and are in emotional denial of their losses. Trading book rules are designed to protect traders who are gamblers. People who like trading because they like gambling are always going to be terrible at it. For these people, the trading books could be greatly shortened to the message: “Don’t trade. You are really bad at this. So just don’t do it.” I don’t actually have a gambler’s mentality. I make different emotional mistakes. So, imposing trading book rules on me is a terrible misfit. p47
In those early days, I wasn’t setting stops at levels that made sense based on the underlying hypothesis for the trade; I was setting stops based on my pain threshold, and the market doesn’t care about your pain. I learned from that mistake. When I get out of a trade now, it is because I was wrong. p48
I’d like to do a better job of monetizing other people’s irrational euphoria. p49
There are various gauges. A simple one is just price action. If it trades like a bull market, it’s a bull market. Another indication is how passionately people defend things that make no sense. For example, some people believe that Barack Obama is not a U.S. citizen. The point is that beliefs that are completely invulnerable to evidence and passionately defended can be quite durable. It has nothing to do with the fundamental logic. p49
Gold is special, magical, and great. It’s not. But if people believe it, they buy it. And if they buy it, it goes up. That’s why there’s a bull market. You can’t go to a meeting without someone saying, “What do you think about gold?” p50
The thing about gold is that if you told me gold has a price of $100, that’s fine. If you told me it’s $10,000, that’s fine as well. It can be any price. Gold is worth exactly what people think it’s worth. p50
I think the natural way to trade a market that is in a bubble is from the long side, not the short side. You want to be long the exponential upmove without taking on the gap risk of a collapse. Therefore options provide a good way of doing this type of trade. p51
My natural trade time horizon is one to three months, but that doesn’t mean it would be right for you. p53
Are there traits that determine who will be a successful trader? Perseverance and the emotional resilience to keep coming back are critical because as a trader you get beaten up horribly. p54
[On how to pick a stop:] First, you decide where you are wrong. That determines where the stop level should be. Then you work out how much you are willing to lose on the idea. Last, you divide the amount you’re willing to lose by the per-contract loss to the stop point, and that determines your position size. p57
O’Shea believes that the best way to trade a market bubble is to participate on the long side to profit from the excessive euphoria, not to try to pick a top, which is nearly impossible and an approach vulnerable to large losses if one is early. The bubble cycle is easier to trade from the long side because the uptrend in a bubble is often relatively smooth, while the downtrend after the bubble bursts tends to be highly erratic. p59
O’Shea would never be outright long in a bubble market, but instead would express a bullish posture through a position such as a long call, a trade in which the maximum risk is defined by the premium paid for the option. Low volatility bubble markets are especially well attuned to being traded via long calls. p60
One common theme that seems to underlie almost all the trades that O’Shea discussed in this chapter is that they are structured to be right skewed—that is, the maximum loss is limited, but the upside is open-ended. p61
Chapter 2: Ray Dalio
The flagship fund uses the relatively rare combination of a fundamentally based systematic approach. (Most hedge funds that are fundamentally based use a discretionary approach, and most hedge funds that use systematic approaches base them on technical input.) p63
The fund combines two factors—growth and inflation—and two states—increasing and decreasing—and comes up with four conditions: 1. Growth increasing 2. Growth decreasing 3. Inflation increasing 4. Inflation decreasing. This four-part categorization reflects Dalio’s view that changes in expected growth and expected inflation are the dominant reasons that some asset classes do well when others do poorly. p65
Create a culture in which it is okay to fail but unacceptable not to identify, analyze, and learn from mistakes. p66
Dalio leaves little doubt about his views on the topic of openness and honesty. For example, Management Rule #11 in Principles states: “Never say anything about a person you wouldn’t say to him directly. If you do, you’re a slimy weasel.” p67
Dalio tends to think in terms of interconnections rather than linearly, which can lead to rambling answers, as he readily acknowledges, “I see things in complex ways, and I have a problem communicating my way of seeing to others.” p67
There are ways to structure your trades so that you can produce a whole bunch of uncorrelated bets. You have to start with your goal. My goal is that I want to trade more than 15 uncorrelated assets. You are just telling me your problem, and it’s not an insurmountable problem. I strive for approximately 100 different return streams that are roughly uncorrelated to each other. There are cross-correlations that enter into it, so the number works out to be less than 100, but it is well over 15. p72
People think that a thing called correlation exists. That’s wrong. What is really happening is that each market is behaving logically based on its own determinants, and as the nature of those determinants changes, what we call correlation changes. For example, when economic growth expectations are volatile, stocks and bonds will be negatively correlated because if growth slows, it will cause both stock prices and interest rates to decline. However, in an environment where inflation expectations are volatile, stocks and bonds will be positively correlated because interest rates will go up with higher inflation, which is detrimental to both bonds and stocks. So both relationships are totally logical, even though they are exact opposites of each other. If you try to represent the stock/bond relationship with one correlation statistic, it denies the causality of the correlation. Correlation is just the word people use to take an average of how two prices have behaved together. p72
When I am setting up my trading bets, I am not looking at correlation; I am looking at whether the drivers are different. I am choosing 15 or more assets that behave differently for logical reasons. p73
Principles determine what we do under different circumstances. In other words, we make decisions about the criteria we use to make decisions. We don’t make decisions about individual positions. p74
I don’t believe in reducing exposures when you have a losing position. I want to be clear about that. The only pertinent question is whether my being in a losing position is a statistically meaningful indicator of what the subsequent price movement will be. And it is not. For that reason, I don’t alter positions because they are losing. p76
Beginning around 1980, I developed a discipline that whenever I put on a trade, I would write down the reasons on a pad. When I liquidated the trade, I would look at what actually happened and compare it with my reasoning and expectations when I put on the trade. Learning solely from actual experience, however, is inadequate because it takes too much time to get a representative sample to determine whether a decision rule works. I discovered that I could backtest the criteria that I wrote down to get a good perspective of how they would have performed and to refine them. The next step was to define decision rules based on the criteria. I required the decision rules to be logically based and was careful to avoid data mining. p77
The truth about hedge funds is that much of what is packaged as alpha is really beta sold at alpha prices. The average hedge fund is about 70 percent correlated with stocks. p81
Why are most hedge funds so skewed toward strategies that do well in good times? I think it is human nature for people to choose strategies that worked well during the recent past, which implies a long bias. p81
The biggest mistake investors make is to believe that what happened in the recent past is likely to persist. They assume that something that was a good investment in the recent past is still a good investment. p85
Typically, high past returns simply imply that an asset has become more expensive and is a poorer, not better, investment p85
Investors are often baffled when markets respond to news events in counterintuitive fashion. p87
Real per capita GDP in the United States has increased at an average rate of near 2 percent over the past 100 years as a result of productivity gains, but has fluctuated widely around this trend based on the prevailing long-term and business cycles. p88
Chapter 3: Larry Benedict
Benedict is essentially a mean reversion trader—that is, he will be a seller into short-term upswings and a buyer into short-term declines. p95
It’s not about making a lot of money. Of course, it’s important to have return, but for Larry it’s more important not to lose money. He knows that if he can manage the risk, he will make money. p95
Fundamental managers say, ‘I’m going to buy this stock for the next six months because I think this will happen.’ Technical managers say, ‘If the stock goes here I will buy, and if it goes here I will sell.’ p96
You’re going to get fired every day. Just come back the next day. […] We go back to the office after the first day of work, and he says, “My account is off by 20 lots. Get the hell out of my office, and don’t come back. You’re fired.” What did you do? I came in to the office the next morning, tail between my legs, and sat down at the desk. He didn’t say anything and acted as if everything was normal. And then two days later he says, “You don’t know what you’re doing. You’re fired.” p99
As the specialist you provide liquidity. You also have an inherent advantage in trading because you see the order book. p103
If I was a seller of calls, I would buy futures against it. I would keep the book close to delta neutral. So you were essentially making the bid/ask spread and laying off the risk. p103
I trade a lot of mean reversion. One of the things I do is three days up, three days down. If the market is up three days in a row, I want to start getting short, and if it’s down three days in a row, I want to start getting long. p104
I have a broad global macro view, but I express it in a day or a few days in duration. p105
I let the market dictate to me how I should be trading, not my macro views of what I think the market will do. p105
In October 2008, I did something I normally don’t do—I let a market assumption that I thought was right, but ultimately proved to be wrong, influence my trading. At the time, with the U.S. in a financial meltdown, I thought that Japan would dramatically outperform the U.S. because it had already had its banking crisis. My mistake was that every hedge fund manager in the world was long Japan, Hong Kong, and China, and those markets went down more than the U.S. because they had to liquidate their positions. p105
The growing influence of high-frequency trading has changed the behavior of the market and has made it more difficult for someone like me who is a pure tape reader looking for clues in the market action. p105
I learned that you can’t be emotional in this business. It’s a business and nothing else. p109
One of the hard things about managing client money is that although I am very patient, the clients aren’t very patient. One of my problems is that I want to make everyone happy. p110
As soon as I put on a position, I immediately start looking for a trade that would be the best offsetting hedge. Say I go long the S&P, I would be looking at my 10 screens for the best hedge. It might be selling another index, or selling a deep-in-the-money call, or selling an out-of-the-money call, or buying bonds (when it is inversely correlated), or taking an inversely correlated position in the euro. And sometimes, I will not put on any offsetting trade, but just use a stop on the position. p111
I came in long because today was option expiration day, and 85 percent of the time, they mark up the market on the morning of option expiration day. It has been going on for 25 years. p112
The essence of Benedict’s approach is that he looks at markets in context of the price action in other markets, rather than in isolation. Markets are correlated, but these correlations come and go and can change radically over time. There are times when the S&P and T-bonds will move in the same direction, and times when they will move in opposite directions. There are times when the S&P 500 will follow crude oil prices, and times when the stock market ignores crude oil prices. p113
The relevant lesson for traders is that the price action of other markets can contain useful information. p114
The biggest principle Larry pushes is that you are not a trader; you are a risk manager. p115
A short straddle is a combined short call and short put position. A short straddle will make money (all or part of the premium received for selling the options) if the market trades within a moderate price range. It is, effectively, a bet that the market will not be volatile. If the market, however, witnesses a sharp swing in either direction, the short straddle position can be subject to large losses. The maximum profit on a short straddle position is limited to the premium received. The maximum loss, however, is not merely unlimited, but accelerates exponentially if there is a large price move in either direction. p115
Chapter 4: Scott Ramsey
Ramsey begins by establishing a broad fundamental macro view that determines his directional bias in each market. Once this bias is established, he will seek to go short the weakest market in a sector if he is bearish or long the strongest market if he is bullish, using technical analysis to time trade entry and position adjustments. Ramsey will score his best returns when he gets the fundamentals right, but even when he is wrong, his rigorous risk control keeps losses relatively small. p116
Our economics professor suggested that we subscribe to the Wall Street Journal, which I did. p118
I made every rookie mistake in the books. Rather than taking the easy route and trading with the trend, I was trying to pick tops and bottoms, and I sat with losers and took small profits. p120
The evolution of a trader is when you start letting your money work for you and increasing your size. p123
I learned the value of the classic “buy the rumor, sell the facts” kind of trading because it put you on the opposite side of retail buying. p124
While I try to keep emotions out of my trading, which is the real challenge, I try to imagine how the guy who took the textbook trade and is losing money feels. Where will he finally capitulate? Is this the start of a big move the other way? p124
It became a matter of not only looking at prices on a chart, but also thinking about why prices were where they were, how people were positioned, and the psychology of the market. You have to try to understand what people are thinking. I began to look at the market from the perspective of other traders. What would I be feeling if I were short bonds, thinking I had this great trade, and seeing the market go against me? p126
The reality is that I’m not being paid to be right; I am being paid to make money. You have to have a degree of flexibility. Whenever I talk to investors, I make it clear to them that whatever I say today about the markets may or may not reflect the positions I have tomorrow or the next day. p126
I use the Relative Strength Index (RSI), but not as an overbought/oversold indicator; instead, I look for divergences between the RSI and price. I look at the 200-day moving average and Fibonacci retracements. If the market sells off to the 200-day moving average and I’m short, I may be inclined to take the money off the table and watch how the market reacts. I particularly like to get combined signals, such as the price approaching the 200-day moving average and a 50 percent retracement. p130
There are such great benefits to trading in exchange-traded markets or the interbank market that giving up those benefits is just not worth it. Lower liquidity doesn’t work with my style. I need to be able to stop myself out in 10 basis points if I want to. p131
Once he has established a firm fundamental opinion, Ramsey utilizes technical analysis to confirm his anticipated scenario. p136
Ramsey will always buy the strongest market in a sector for long positions and sell the weakest market in a sector for short positions. Many novice traders make the error of doing the exact opposite. They will buy the laggards in a sector on the typically mistaken assumption that those markets haven’t yet made their move and therefore provide more potential and less risk. p137
Although the use of a 0.1 percent stop point from entry is probably too extreme (or perhaps even inadvisable) for most traders to adopt, the general concept of using a relatively close stop on new trades and allowing a wider stop after a profit margin has been created is an effective risk management approach that could work well for many traders. p138
Chapter 5: Jaffray Woodriff
Woodriff and Jordan had never bothered to sign a partnership agreement. Up to that point, they had split the firm’s small net earnings. With his systems beginning to generate significant profits, Woodriff thought it was time to formalize their business relationship with a written agreement. This decision would probably not have been problematic in itself, but the terms that Woodriff proposed so angered Jordan that he broke off the partnership and filed a lawsuit against Woodriff. Woodriff’s relationship with Blue Ridge ended shortly thereafter, and the firm was closed down a few months later. p142
When I was 18, I started checking the prices of the stocks in my trust account in the daily paper. I checked it for a few days, and it bored me because the prices didn’t change much from day to day. But I noticed that the option prices moved much more, that is, in percentage terms. p148
That summer I tried daytrading. I got a whole real-time quote setup. My plan was to spend the summer staring at the trading screen. After about the third day of doing that, I realized, This isn’t me. This just doesn’t work for me. p150
People ask me all the time if I had any mentors. The truth is that I am self-taught. I have thought through this so many ways, and there is no other way to put it. But if you ask me what manager I set out to emulate, it’s clearly Paul Tudor Jones. p152
I started looking at trend-following models, which was kind of interesting, but I thought there were already lots of successful trend followers out there, and I didn’t want to compete with them. I wanted to do something different. p153
My contrarian nature was opposed to doing what other people were doing. I also believed that if everyone thought trend following worked, it would start to work less well. I was never going to be a trend follower. That was clear. Mean reversion was more interesting. But I didn’t like mean reversion either. p153
I discovered that it was much better to use multiple models than a single best model. p154
I blindly search through the data. It’s nice that people want hypotheses that make sense. But I thought that was very limiting. I want to be able to search the rest of the stuff. I want to automate that process. p161
Luckily for me, my intuition that there was an edge available in a nontrend following, noncountertrend type of model was correct. My intuition was that there should be other emergent patterns in the price data that are more complex than trend following. p162
A lot of people think they are okay because they use in-sample data for training and out-of-sample data for testing. Then they sort the models based on how they performed on the in-sample data and choose the best ones to test on the out-of-sample data. The human tendency is to take the models that continue to do well in the out-of-sample data and choose those models for trading. That type of process simply turns the out-of-sample data into part of the training data because it cherry-picks the models that did best in the out-of-sample period. It is one of the most common errors people make and one of the reasons why data mining as it is typically applied yields terrible results. p163
To avoid hindsight bias error in developing trading systems, the available past data is segmented into seen data (i.e., “in-sample”) that is used for system development and unseen data (i.e., “out-of-sample”) that is used for system testing. Any results on the in-sample data are ignored because they are hindsight-biased. Although segmenting the data to reserve unseen data for testing is a necessary condition to avoid misleading results, it is not a sufficient condition as Woodriff goes on to explain.
Not only does the software fail to guide the user in doing data mining correctly, it actually leads users in the wrong direction because it allows them to generate bogus evidence to support their pet theories. p164
One of the changes we have made from the early years, which has greatly expanded capacity, is to execute trades throughout the trading session as opposed to only on the opening. p164
there is a strong pattern for our edge to be greater in more liquid markets. So besides increasing capacity, the shift to allocating a greater percentage to more liquid markets has also improved performance. p164
The OTC markets are very often used to take advantage of clients who are “sophisticated” in the legal definition, but are naive in practice. The OTC markets have been built to maximize asymmetries of information and are an example of how markets should not operate. p166
How would you summarize the trading rules you live by? Look where others don’t. Adjust position sizes to overall risk to target a particular volatility. Pay careful attention to transaction costs. p167
When I was in my teens, my highly insightful father was somehow able to instill in me the discipline of objectively evaluating your own progress. That lesson, more than anything else, has been critical to my success. p167
Systems that work well across many markets are more likely to continue to work in actual trading than systems that do well in specific markets. The lesson is: Design systems that work broadly rather than market-specific systems. p167
Data mining refers to the process of using computers to analyze large amounts of data to discover patterns in the data. Although data mining techniques can uncover patterns in data that would be impossible for humans to find empirically or by prior hypotheses, it can also identify meaningless patterns that are nothing more than chance occurrences or the product of flaws in the analytical process. When searching very large numbers of combinations of past price data for patterns, it is easy to come up with many patterns that worked well in the past simply by chance, but have no predictive value. This common pitfall of applying data mining to price data is the reason why the term often has derogatory connotations in reference to trading systems. p169
Part Two: Multistrategy Players
Chapter 6: Edward Thorp
You can’t prove a negative. I can’t prove it doesn’t work. All I can say is that I did not see enough substance there to pursue it. I didn’t want to take time to try things unless I thought they were pretty good. p198
I realized that the pricing of warrants should be mathematizable. The huge amount of variables that might apply to trying to forecast stock prices were almost all eliminated if you focused on warrants instead. I wrote down a short list of the factors that I thought should determine the warrant price. These included the price of the underlying stock, the strike price, the volatility, the time to expiration, and the interest rate level. There were all the same factors that everyone today agrees determine option prices. p199
Historically ideas don’t just appear in one place; they tend to appear in several places at almost the same time. p200
Warrants with less than two years to run typically traded at premiums that were too high. The typical trade we did was to short the warrant and hedge it by buying the stock. p200
We started out with a static hedge and then decided that dynamic delta hedging was better. p200
Delta hedging is an options trading strategy that aims to reduce, or hedge, the directional risk associated with price movements in the underlying asset. The approach uses options to offset the risk to either a single other option holding or an entire portfolio of holdings. The investor tries to reach a delta neutral state and not have a directional bias on the hedge. https://www.investopedia.com/terms/d/deltahedging.asp
Kassouf thought he could tell something about the direction of stock prices and would sometimes take a fundamental view on a stock, whereas I was afraid to do that because I didn’t believe I had any forecasting power. I thought we should always hedge to be delta neutral. p200
The Kelly criterion is the fraction of capital to wager to maximize compounded growth of capital. Even when there is an edge, beyond some threshold, larger bets will result in lower compounded return because of the adverse impact of volatility. The Kelly criterion defines this threshold. p203
He told me to look at Kelly’s paper in deciding how much to bet because in favorable situations, you will want to bet more than in unfavorable situations. […] The Kelly criterion of what fraction of your capital to bet seemed like the best strategy over the long run. p204
If you can calculate the probability of winning on each bet or trade and the ratio of the average win to average loss, then the Kelly criterion will give you the optimal fraction to bet so that your long-term growth rate is maximized. p205
There is an important distinction between trading and playing a game such as blackjack. In blackjack, theoretically, you can know the precise probabilities, but in trading, the probability of winning is always an estimate—and often a very rough one. p206
Say I am playing casino blackjack, and I know what the odds are. Do I bet full Kelly? Probably not quite. Why? Because sometimes the dealer will cheat me. So the probabilities are a little different from what I calculated because there may be something else going on in the game that is outside my calculations. p207
One of the researchers looked at stocks that were most up and stocks that were most down during the recent past. He found that the stocks that were most up tended to underperform the market in the next period, while the stocks that were most down tended to outperform the market. That finding led to a strategy of buying a diversified portfolio of the most down stocks and selling a diversified portfolio of the most up stocks. We called that strategy MUD for most up, most down. p208
There was effectively a U-shape curve that favored buying stocks with no dividends and high dividends and selling stocks with low dividends. p209
We primarily traded Japanese warrants, which was very profitable until the dealers began dramatically widening the spreads on us. The trades might have had a 30 percent profit potential, but the bid/ask spreads became so wide that the transaction costs were 15 percent. p215
Since correlations between markets change so radically over time, even changing sign, how long of a lookback period did you use? We found that 60 days was about best. If you use too short of a window, you get a lot of noise; if you use too long of window, you get a lot of old information that isn’t relevant. p220
My view on trend following was that I could never be sure that I had an edge, so I wanted a safety mechanism. Whereas for a strategy like convertible arbitrage, I had a high degree of confidence as to the payoff probabilities p221
Based on our daily success in Princeton Newport Partners, the question wasn’t “Is the market efficient?” but rather the appropriate questions were “How inefficient is the market?” and “How can we exploit the inefficiencies?” The claim of market efficiency, which implies that no market edge is possible, is a hollow statement because you can’t prove a negative. But you can disprove market efficiency if there are people who have a demonstrable edge. There is a market inefficiency if there is a participant who can generate excess risk-adjusted returns that can be logically explained in a way that is difficult to rebut. Convertible arbitrage is a good example. You can lay out exactly how it works, why it works, and approximately how much return you expect to get. p224
Beating roulette was once considered a mathematical impossibility. By approaching the problem in a completely different way—focusing on predicting the most likely termination zone for the ball—Thorp was able to create a giant 44 percent edge. The broad lesson is that sometimes what seems impossible is entirely possible if approached from a completely different perspective. p225
Delta neutral hedging means that purchases and sales of the underlying stock are used to keep the combined position balanced so that it is approximately unaffected by small price changes in either direction. p230
Chapter 7: Jamie Mai
Cornwall’s strategies range from thematic fundamental trades to trades that seek to profit from esoteric market inefficiencies. The one unifying characteristic virtually all of Cornwall’s strategies share is that they are structured and implemented as highly asymmetric, positive skew trades—that is, trades in which the upside potential far exceeds the downside risk. p232
There is truth in the adage that accounting is the language of business. p237
finding answers is much easier when you know in advance what the questions are. Understanding what the right questions are can be deceptively difficult, but once you do, the rest is straightforward deductive analysis to determine whether the hypothesis is right. Usually it isn’t, in which case you throw it into the waste bin with the 98 percent of other hypotheses that never amount to anything. p238
As a general observation, markets tend to overdiscount the uncertainty related to identified risks. Conversely, markets tend to underdiscount risks that have not yet been expressly identified. p240
It seemed quite clear to us that substantial new information was likely to come out relatively soon that would either validate the rating agencies’ concerns about the litigation, in which case, the stock price would sell off substantially, or the reverse, in which case, the stock would appreciate substantially. Thus there was an above-average probability of a large price move in one direction or the other. We had already seen cases where the option market assigned normal probability distributions to situations that clearly had bimodal outcomes. p241
the first thing we checked was whether the Altria options still assumed a normal probability distribution, despite the presence of a bimodal event. Sure enough, the Altria option prices still implied a normal distribution, which meant the out-of-the-money options were way too cheap. p241
By being long 10-year call options, we are taking exposure on the risk-free rate implicit in the option pricing models. If interest rates go up, the value of the options can go up dramatically. p243
In a worst of option, you pay a single premium for a basket of options. In the trade we did, there were puts on two crosses in the basket: euro versus Australian dollar (EUR/AUD) and euro versus Swiss franc (EUR/CHF). Worst of option structures are cheaper because the payout is determined by whichever option performs more poorly from the buyer’s perspective. As long as one of the options expires out-of-the-money, the option buyer will lose the entire premium. The expected correlation between the different options comprising a worst of basket is therefore a relevant price input, in addition to the standard inputs that determine option prices. p245
When you own options, you’re always fighting against the time decay. Figuring out how to make the option premium cheaper is one way of mitigating that decay. p249
The broader principle is that the explicit and implicit assumptions that go into option pricing models often diverge from the underlying reality. Looking for those divergences can be a very profitable exercise because you can wait and do nothing until you see a probability that is wildly mispriced. Option math works a lot better over short intervals. Once you extend the time horizon, all sorts of exogenous variables are introduced that can throw a wrench into the option-pricing model. p250
To varying degrees, all of the trades we’ve discussed so far have reflected situations where we did not have a high level of conviction ourselves about the outcome we were seeking to make money on. Instead, we had conviction that the odds were substantially mispriced, providing us positive expected value, even though we might not have had a strong view about the direction of the underlying market. p251
Whereas the classic value investor achieves capital preservation by taking risks only when he is confident that he won’t lose a meaningful amount of money, we think about risk more probabilistically. p251
It is hard to see any circumstances in which we would go outright short a stock—your gain is limited, your loss is unlimited, and your exposure grows as you are wrong. Instead, we bought out-of-the-money puts on some of the dry bulk ship operators. p264
Find mispricings in a theoretically priced world. Mai seeks to identify trade opportunities that arise because prices, particularly for derivatives, are based on one of a number of standard pricing assumptions that may be entirely inappropriate based on the specific circumstances applicable to the given market. When these assumptions are unwarranted, they create mispricings and trading opportunities. p164
As a general rule of thumb, Cornwall requires that the estimated gain if the trade succeeds multiplied by the probability of a positive outcome must be at least twice as large as the estimated loss if the trade fails multiplied by the probability of a negative outcome. Of course, these gain-and-loss amounts and their respective probabilities must be based on subjective estimates. p264
Implement trades asymmetrically. Mai structures trades so that the downside is severely limited, while the upside is open-ended. One common way of achieving this type of return/risk profile is by being a buyer of options (of course, only at those times when a mispricing is identified). p265
Wait for high-conviction trades. Mai is perfectly content to stay on the sidelines and do absolutely nothing until there is a trade opportunity that meets his guidelines. p265
Options are priced based on the assumption of a normal distribution, which effectively implies that future prices near the current level are most probable and that probabilities drop steeply for prices further removed from current levels. In some instances, however, large price moves are much more likely than implied by the normal distribution. Consider, for example, Mai’s comments regarding Capital One: “The odds that the stock would still be near $30 in two years seemed vanishingly small. Either the company would be vindicated or it would go under.” In such circumstances, options priced in line with option-pricing models will be severely mispriced; specifically, out-of-the-money options will be priced far too cheaply. These mispricings can create profit opportunities. If the probability of a large price move is much greater than normal, as was the case with Capital One, then out-of-the-money options can provide a significant probability of gain, while being priced consistent with long-shot probabilities. p265
Another implication of the normal distribution assumption is that it is always equally likely for prices to go up X percent as down X percent. Although this assumption may often be a reasonable approximation, there are times when it is far more likely for a market to go up by a given percentage than down by the same amount or vice versa. p266
Good traders get out of a position when they realize they have made a mistake. Great traders are capable of taking the opposite position when they realize their original concept was dead wrong. p269
Chapter 8: Michael Platt
Markets may initially trend for fundamental reasons, but prices overshoot by ludicrous amounts. At some point, prices go up today simply because they went up yesterday. p275
Can’t equities be quantitative? We do have a small program, which is a systematic program trading single stocks. It is basically an intelligent market maker. It models the price of every stock, using all of the other stock prices as the input. It looks for divergence. If, for example, Fidelity decides to liquidate a given stock, their selling will probably push the stock out of line with the basket of stocks. We will buy the stock and sell the basket, and we will do the same thing with other stocks. So it is basically a statistical arbitrage approach. Yes it is. We have done a phenomenal amount of work on quantitatively selecting single stocks. p279
I look for the type of guy in London who gets up at seven o’clock on Sunday morning when his kids are still in bed, and logs onto a poker site so that he can pick off the U.S. drunks coming home on Saturday night. I hired a guy like that. He usually clears 5 or 10 grand every Sunday morning before breakfast taking out the drunks playing poker because they’re not very good at it, but their confidence has gone up a lot. That’s the type of guy you want—someone who understands an edge. Analysts, on the other hand, don’t think about anything else other than how smart they are. p286
I look at each trade in my book every day and ask myself the question, “Would I enter this trade today at this price?” If the answer is “no,” then the trade is gone. p289
Part Three: Equity Traders
Chapter 9: Steve Clark
I am not a big believer in chart analysis. It is extremely appealing to think that you can take a data set from the past and predict what will happen in the future. It is very attractive because as a human being you are always looking for certainty. You can use charts to give you a plus or minus toward your view, but you can never start with the chart. To say that you can predict the future from past data is patently untrue. You can talk about percentage probabilities of what might happen next, but you can’t go any further than that. p302
You need to be willing to call people to get information. You need to ask the next question. You never know what you will find out that will be useful. For example, there was a merger deal that had to be approved by the FCC, and nothing was happening. In the meantime, the spread kept widening and widening, and we were getting more concerned about the position. We kept calling the FCC over and over again until one time, by luck, we got the secretary of one of the commissioners, who said, “Oh, that; it’s on his desk ready to be signed, but he is away fishing this week.” p320
Nearly all the successful traders I have known are one-trick ponies. They do one thing, and they do it very well. When they stray from that single focus, it often ends in disaster. p322
Never stop asking questions. Speak to as many people as you can. Research every opposing opinion. p326
The market is not about facts; it’s all about people’s opinions and positions. Consequently, anything can be at any price, any time. p326
Traders focus almost entirely on where to enter a trade. In reality, the entry size is often more important than the entry price because if the size is too large, a trader will be more likely to exit a good trade on a meaningless adverse price move. The larger the position, the greater the danger that trading decisions will be driven by fear rather than by judgment and experience. p327
Chapter 10: Martin Taylor
By buying options, the downside is limited and the upside is unlimited. That sounded good to me. p335
In an emerging market, the smart money is domestic, not international. p343
There are multiple requirements for a trade to take place. Do we like the company? Is it cheap? Does it generate cash flow? Do we trust the management? Do I have confidence in my projections? Is the macro outlook favorable? If a company meets all the criteria, then the next step is determining the appropriate position size. Given the degree of company and country risk, what do I think is the appropriate size position? If it is something very risky, a large position for me might be 1 percent to 5 percent of the portfolio. If it is a lower risk position in which I have very high confidence, the position could be as large as 20 percent of the portfolio p359
Say I really like a company and want to put 10 percent of the portfolio in it. If the stock is extremely overbought, then I might start with only a 1 percent position. If the stock just keeps on going up, then I am happy that I bought at least some. I will also be more willing to buy more because I bought part of the position at a lower price. Whereas, if I didn’t initially buy anything because the stock was overbought, I would then never buy any of it, which would be a dreadful mistake. p361
RSI doesn’t work as an overbought indicator because stocks can remain overbought for a very long time. But a stock being extremely oversold is usually an acute phenomenon that lasts for only a few weeks. […] If the RSI is extremely oversold, I will then look at the fundamentals to see if whatever has caused the stock to be sold off that sharply is already discounted by the price. If it is, then I will buy. p362
You have to be an expert in what you invest in. You need to understand why you are invested. If you don’t understand why you are in a trade, you won’t understand when it is the right time to sell, which means you will only sell when the price action scares you. Most of the time when price action scares you, it is a buying opportunity, not a sell indicator. p363
Taylor believes the best opportunities are those where you can identify a potential trend that the market does not appreciate because it is extrapolating history instead of looking forward. p366
Chapter 11: Tom Claugus
Just because you made money doesn’t mean you were right, and just because you lost money doesn’t mean you were wrong. It is all a matter of probabilities. If you take a bet that has an 80 percent probability of winning, and you lose, it doesn’t mean it was a wrong choice. p384
Whether a trading decision was right or wrong is not a matter of whether you won or lost, but rather whether you would make the same decision all over again if faced with the same facts. p384
There are always some stocks that are going down. The interesting thing is that shorts are actually easier to find than longs. It is easier to spot a broken company than a good company. It is easier to identify bad management than good management. p385
Trading is a matter of probabilities. Any trading strategy, no matter how effective, will be wrong a certain percentage of the time. Traders often confuse the concepts of winning and losing traders with good and bad trades. A good trade can lose money, and a bad trade can make money. p392
Chapter 12: Joe Vidich
Pricey stocks are always 30 percent pricier than they should be because people are willing to own them at 30 percent above what they should own them at. A good growth stock is always overvalued, and a lousy company is always undervalued. That is the danger of buying value stocks. Until you get the turn where the market recognizes an improvement in the business model, they are always going to be undervalued. If you are going to short a growth stock because it is 30 percent overvalued, it is going to grow for the next five years always being 30 percent overvalued—until it finally breaks, and by that time, you are probably not going to be there. p399
One thing that I learned as a trader is that if the market [Dow] moves up 400 points in the morning, you usually don’t go much further. Every time there was a big rally, we would put our shorts on, and on a selloff, we would cover the shorts. p401
When P/E multiples get to 50, 60, or 70, you are in the realm of concept stocks. p402
He didn’t have a good reason. He wanted to sell to justify his idea and lock in gains on his recommendation. But that is his own world. It has nothing to do with where the stock is going. p407
Getting out sometimes right before a stock turns is the price you pay to keep your losses under control. p410
Don’t try to be 100 percent right,” he says. Instead of making an all-or-nothing decision, when Vidich is faced with a losing position, he will often begin by liquidating part of the position. Taking a partial loss is much easier than liquidating the entire position. It allows the trader to act rather than procrastinate. If the position continues to move against Vidich, he will liquidate some more. In this way, a losing position is gradually reduced (and eventually entirely liquidated if it doesn’t turn around), mitigating the damage. The next time you are undecided between liquidating a losing position and gritting your teeth and riding it out, remember that there is a third alternative: partial liquidation. p411
Don’t make trading decisions based on where you bought (or sold) a stock. The market doesn’t care where you entered your position. p412
Chapter 13: Kevin Daly
He runs the fund as a solo operation from his home office. Daly is quite happy with this simple structure, and he is downright reluctant to put any effort into raising assets for his fund, as it might necessitate expanding beyond a one-man business. He is loath to complicate what he considers the perfect current arrangement. Daly is a man whose work and hobby are the same. If Daly were retired, I believe his life would be totally unchanged. p414
I have tried to do that type of investing throughout the fund’s life—looking for companies that are somewhat obscure, not covered by Wall Street, or have a niche business inside the company that no one is focusing on. p419
Treat investments as a business, not as a gamble. p428
Chapter 14: Jimmy Balodimas
I always take some money off the table when the market is in my favor. Regardless of whether it is a lot or a little bit. p433
That is a habit I have had since day one. I always take money off the table when it’s in my favor. Always, always, always. p433
Buy companies when they are out of favor and sell companies when everybody loves them. p439
I don’t mind if I miss a move down because the market will always have a move that I can re-enter if I really feel like it. I am never afraid of missing anything. The market is always providing opportunities. p442
All I think about is making money; not being right. p442
I am a natural short because there is a hype, a marketing campaign that is always at work to get people to be long stock, and it’s unnatural. It’s an energy that can’t always be maintained. p443
Although cloud computing is a great story, it is overplayed. Current prices way outstrip any growth potential for the next few years. Analysts are talking about 2015 numbers. It’s ridiculous. No one knows what is going to happen four years out. It’s difficult enough to forecast the next quarter. The rhetoric buildup is always a similar pattern. p445
I really can’t imagine myself sitting in front of the screens for the rest of my life. p452
Chapter 15: Joel Greenblatt
The way I looked at the risk/reward of the stock was that it was a binary situation: The stock would go down $80 if Wells Fargo went out of business and up $80 if it didn’t. But by buying LEAPS with more than two years until expiration instead of the stock, I could turn that 1:1 risk/reward into a 1:5 risk/reward. If the bank survived, the stock should be a double, and I would make five times my money on the options, but if it failed, I would lose only the cost of the options. p466
By using EBIT (which looks at actual operating earnings before interest expenses and taxes) and comparing it to enterprise value, we can calculate the pre-tax earnings yield on the full purchase price of the business (i.e., pre-tax operating earnings relative to the price of equity plus any debt assumed). This allows us to put companies with different levels of debt and different tax rates on equal footing when comparing earnings yields. p467
The power of value investing flies in the face of anything taught in academics. Value is the way stocks are eventually priced. It requires the perspective of patience because the market will eventually gravitate toward value. p470
If I wrote a book about a strategy that worked every month, or even every year, everyone would start using it, and it would stop working. Value investing doesn’t always work. The market doesn’t always agree with you. Over time, value is roughly the way the market prices stocks, but over the short term, which sometimes can be as long as two or three years, there are periods when it doesn’t work. And that is a very good thing. The fact that our value approach doesn’t work over periods of time is precisely the reason why it continues to work over the long term. p471
We wanted to limit our concentration in any specific industry group and limit our exposure in any single stock. We have a widely diversified portfolio. In our small-cap portfolio, the largest long position is around 0.6 percent of equity, and the largest short position is even smaller. p474
Stuff happens. Don’t fall in love with any position. Always keep a large margin of safety, even if you’re playing with house money. Even though the stock was still at a discount to what I thought it was worth right before 9/11, it obviously was a much less attractive value than it was before it had quadrupled, and we probably should have taken some profits. Also, operating leverage works both ways. To quote Howard Marks, “Experience is what you got when you didn’t get what you wanted.” p481
What I would say is that to beat the market, you have to do something different from the market. And if you are going to do something different, sometimes you will underperform significantly. For example, if you are a value investor, there will be times when the market will be responding to factors such as emotion and momentum where a value approach might perform poorly. p486
Greenblatt’s view is that although the market will eventually trade at fair value, a price that would be consistent with the efficient market hypothesis, in the interim, which can sometimes be as long as years, stocks can deviate substantially from their fair value. According to Greenblatt, a more appropriate model is that prices trade around fair value, but broad deviations occur because of wide swings in investor emotions. p493
Sometimes Mr. Market is in such a good mood that he names a price that is much higher than the true worth of the business. On such days, it probably would make sense for you to sell Mr. Market your share of the business. On other days, he is in such a poor mood that he names a very low price for the business. On those days, you may want to take advantage of Mr. Market’s crazy offer to sell you shares at such a low price and to buy Mr. Market’s share of the business. p493
Conclusion
Trading to make money is always a bad idea. Traders should only take a trade when the market provides an opportunity as defined by their own individual strategy. p500
Parabolic price moves in either direction tend to end abruptly and sharply. If you are fortunate enough to be on the right side of a market in which the price move turns near vertical, consider scaling out of the position while the trend is still moving in your direction. If you would be petrified to be on the other side of the market, that is probably a good sign that you should be lightening your position. p504
Although being cognizant of correlation between different markets is crucial to avoiding excessive risk, it is important to understand that correlation measures past price relationships. It is only relevant if there is reason to believe that the past correlation is a reasonable proxy for future correlation. Some market correlations are stable, but others can vary widely and even change sign. p507
Option prices are primarily determined by models that assume that large price movements are unlikely. In circumstances when the fundamentals suggest a significant potential for either a large price gain or a large price loss, option prices often fail to reflect the abnormally large probability of such outsized price movements. p509
Appendix B
Roughly speaking, the option buyer accepts a large probability of a small loss in return for a small probability of a large gain, whereas the option seller accepts a small probability of a large loss in exchange for a large probability of a small gain. p523