Staying humble as a trader


The Intelligent Investor by Ben Graham is a classic investing book that raises a question all traders should ask themselves

Are we being too arrogant in thinking that we can beat the markets in the long term? (i.e. 30+ years)
Is outperformance just a matter of luck?

This seems like a weird question to ask, considering that Bull Markets is all about quantitative trading and investing. But the book does make you think and consider your place in the long history of trading.

Main idea

The book’s main argument is that it’s extremely hard to beat buy and hold in the long run (e.g. 30+ years).
Ben Graham, despite his preaching “value investing”, came to the conclusion in his later years that it’s impossible to beat buy and hold in the long term, in real time.
It’s easy to find strategies that outperformed in the past through backtesting. Unfortunately, all strategies that are “proven” through historical backtesting (or even real time usage) have 1 of 2 problems:

  1. The “proven” strategy was data mined, which means that its outperformance was probably just a fluke. In a world with infinite data and infinite trading strategies, there will always be some that “worked” well in the past simply through luck. E.g. if you have 1 million monkeys playing a slot machine, a few of them will succeed through luck.
  2. If a strategy that outperformed in the past is truly sound, then it will be disseminated and copied very quickly by many other traders. By becoming popular, its ability to outperform will shrink over time. (I don’t entirely agree with this second point, but more on this later).

Recent data on factor investing supports Ben Graham’s second point.

The Intelligent Investor Notes

How to succeed in investing/trading
Graham states that in order for a strategy (discretionary or quantitative) to outperform in the long run:

  1. It must be sound (i.e. cannot be data mined). Data mined strategies will be disproven through the mere passage of time.
  2. It must not be popular among the financial community, otherwise its outperformance would be arbitraged away.

For those who have seen the movie or read the book Moneyball, the often assumed lesson is “to succeed, play asymmetric warfare”. (To makeup for their lack of money, the Oakland A’s had to use statistics to “buy” cheap and good players). But the real lesson of the movie is that “once others start to copy your edge, it’s no longer an edge”. (While the Oakland A’s pioneered the mass usage of statistics in baseball, they didn’t succeed with it once other teams like the Boston Red Sox copied their strategy).
*Private equity has outperformed public markets in the long run. However, Ben Graham in the 1970s believed that this outperformance would wither away someday when private equity became popular. Private equity is extremely crowded nowadays.
Things that worked in the past, stop working
The book describes many cases of strategies and ideas that “worked” (outperformed) in the past, but no longer outperformed when they became disseminated.
It really makes you wonder if you are just repeating history. I.e. a trader in the 1940s who thought “my strategy has been proven over the past 75 years” uses his strategy, and then it stops working. Are you doing the same thing as he did all those years ago?
Valuations
Ben Graham used a valuation method for the U.S. stock market which included interest rates and earnings. It was “proven” with 75 years of historical data from 1875-1949 and accurately picked long term tops and bottoms. The instant he published this strategy in a paper in 1949, the valuation method stopped working.
Dow theory
Another example of things that “worked” well in the past and then stopped working was the Dow Theory.
The Dow Theory performed well from 1897 – early 1960s. It became popular because it was “proven” with historical data. The instant the Dow Theory became popular, it stopped working well.
Value investing
A lot of people reading about Warren Buffett’s value investing think “I can do that too”.
But they don’t bother to actually learn how Buffett did value investing throughout his career. Warren Buffett was a true value investor from 1950-1970. He bought what were called “cigar butt stocks”, which were companies so cheap that if they were liquidiated for their assets, the value of those assets would more than cover the market cap of the company. It’s like paying $2 million in cash to buy a company that has land, cash, factories, and assets worth $3 million. You almost cannot lose money.
How could valuations be so low for Buffett to find these opportunities?
Because the 1950s and early-1960s still stood in the shadow of the Great Depression. Interest in the stock market fell off a cliff after the Great Depression. An entire generation was scarred by the stock market crashes of 1929-1932 and 1937-1938. With not many people paying attention to the stock market (and no internet to make information easily accessible), many stocks were priced at below net asset value.
In markets as in life, many of the best opportunities occur when things are unpopular. By the time something becomes popular, much of the low hanging fruit is already taken.
By the late-1960s, Buffett noted that most of these cigar butt stocks were gone because interest in the stock market was renewed by a 20 year bull market. Hence, Buffett completely pivoted away from his original value investing strategy of buying dirt cheap cigar butt stocks.
Interestingly enough, Ben Graham confirms this in his book.
In 1957, a list was published with nearly 200 stocks whose market cap was lower than the value of their net current assets. By the 1960s, there were no stocks whose market cap was lower than the value of their net current assets.
Beware of gurus and financial salesman
The book Intelligent Investor spends a lot of time warning against trading gurus and financial professionals who try to sell you on their latest strategy that “beats buy and hold”. The book warns that marketers are “selling the dream” because the instant something that “works” is popularized and sold to the masses, it tends to stop working.
Information flow is so fast nowadays with the internet that it doesn’t take very long for something that “works” to be disseminated and stop working.
In pre-internet days you could probably discover a strategy that “works” and use it successfully for 10-20 years before others discovered it and arbitraged your outperformance away. With the internet, you will probably discover a strategy that “works” and within a few years find out that it no longer outperforms.
O’Shaughnessy
The book shares an example regarding Jim O’Shaughnessy, the wealthy founder of an asset management firm that sells investment products centered around their “outperformance strategies” (e.g. factor investing)
Jim was rather obscure until 1996, when he wrote a book titled What Works on Wall Street.
Jim’s backtesting demonstrated that you could earn an average of 18.2% a year from 1954 – 1994 if you bought a basket of 50 stocks with the highest 1 year returns, five straight years of rising earnings, and stock prices less than 1.5x their corporate revenues. O’Shaughnessy even patented this discover (U.S. patent # 5,978, 778) and launched 4 mutual funds to invest in this strategy.
The funds gathered more than $175 million from investors, and promptly stopped working.

The fund’s performance lagged the stock market significantly. Apparently 2 of the funds had to shutdown in 2000 due to poor performance
This supports the author’s point that

In finance, the real money is made by selling advice. It doesn’t come from taking the advice.

Challenge your strongest beliefs
Study after study has shown that stocks yield an average of 6-7% (real) in the long run (e.g. 30 year rolling periods).
The book challenges this belief and states that it may not be true. Yes, the “data” does demonstrate that from 1800 – present, the stock market has yielded an average of 6-7% real over 30 year rolling periods.
However, the “data” may be wrong. Reliable figures before 1871 don’t exist. The “indexes” often used to represent the U.S. stock market’s earliest returns contain as few as 7 stocks! (By 1800, there were more than 300 companies in America). Most of these 300 companies went bankrupt.
*By the 1840s, these indices included a maximum of 7 financial stocks and 27 railroad stocks.
Hence, these indices have “survivorship bias”. Yes, those 7 stocks did return an average of 7% from 1800-1871. But if you bought every U.S. stock in the year 1800 as a real index investor would, your returns would have been much lower because you would need to write off all the bankruptcies.
It’s estimated that the “7% a year figure” before 1871 is off by at least 2%.
There is no guarantee that stocks will outperform bonds and other asset classes over a 30 year period. Just look at Japan from 1990 – present.

Warren Buffett says that “a good, growing company isn’t necessarily a good long term investment if its price is astronomically expensive”. The same can be said about a country. Once a country’s stock market is astronomically expensive, it can keep growing in the long run, but the mean reversion will be so strong as to dampen future returns.
Here’s Japan’s Shiller P/E ratio. Notice how extraordinarily high it was by 1990

A get mega-rich scheme
The book describes a popular get rich scheme from the 1990s. Finance professors Jay Ritter and William Schwert demonstrated that if you spread only $1,000 across every IPO in January 1960, at its offering price, sold out at the end of that month, then invested anew in the next month’s IPOs, your portfolio would be worth $533 decillion by the year 2001.
$533 decillion = $533,000,000,000,000,000,000,000,000,000,000,000
(The logic is that if you invested $2,100 in Microsoft when it went public in 1986, that would be worth $720,000 by 2003).
Of course, such a strategy is disastrous when used in real time.

  1. For this strategy to work, you would need to never miss a single IPO over the decades because most IPOs flop and only a few are real winners.
  2. The high returns on IPOs are captured by the investment banks and their best customers. E.g. an investment bank will give its best customers the exclusive change to invest in XYZ at $20. By the time the stock starts trading at 9:30 am, the stock price could easily be at $30. There is no way for the average investor/trader to profit from this.

*This strategy was very profitable in the dot com bubble. Some people made 1000x their money constantly buying IPO bubble stocks. When the dot-com bubble ended, that strategy collapsed to $0
Don’t put all your eggs in one basket
One of the most widespread how-to-be-successful pieces of advice is “put all your eggs in one basket and then watch that basket”.
In essence, do not diversify. Focus.
This is true in a way. All the extremely successful people in the world focused. Bill Gates focused on Microsoft, Mark Zuckerberg focused on Facebook, Sam Walton focused on Wal-Mart.
So if you want to become extremely successful, you need to focus. You cannot diversify.
However, “put all of your eggs in one basket” is bad advice for people as an aggregate.
The problem with any how-to-be-successful advice is that you always hear about the successes and you never hear about the failures. This gives you a false impression as to how good certain advice really is.
If you put all of your eggs in one basket, some people will become extremely successful. These people will be lauded, praised, and broadcasted in the media.
But you will never hear about the 99% of people who put all their eggs in one basket and failed spectacularly.
Here’s an example that the book describes.

  1. In 1980s, the average Forbes 400 member had a net worth of $230 million.
  2. To make it into the 2002 Forbes 400, the average 1982 member needed to only earn 4.5% a year on his wealth. In other words, anyone could have done this if they just stuck their money in a bank account and earned interest or bought and held the S&P 500.

Do you know how many of the 1982 members were still on the list by 2002?
When you don’t include the deaths, only 16%.
Putting all of their eggs in one basket made them extraordinarily successful. But most people don’t stay on top also because they put all their eggs in one basket. Rise by focus, die by focus.
All it took was 20 years. 20 years, and clearly “focus” did not help that group. All of these spectacular failures from the 1982 group seemed inconceivable at the time. E.g. if you tell today’s billionaires “stop whatever you’re doing, sell your company, and just buy and hold the S&P 500”, you would be laughed out the door. A billionaire wants to have $10 billion. But for this group as an aggregate, that is exactly the best advice.
More data here http://www.aei.org/publication/churnin-and-turnin-high-turnover-in-forbes-400/print/
Everyone tells you that if you want to be really successful in life, go and start a business. Only business owners are billionaires.
But did you know that 80-90% of businesses fail? Yes, some will succeed spectacularly. But most will fail. As an aggregate, the average business owner would be better off just working at a 9-5 job and investing his savings into the S&P 500 for the long run.
This train of thought struck me really hard when I read it. I have a family friend who came from nothing and became a billionaire. They grew exponentially by investing and focusing on real estate. The key to real estate is leverage. And over the past 4 years, their real estate empire has crumbed. They are now bankrupt.

  1. Went from zero to $1 billion through leveraged real estate
  2. Went from $1 billion to zero through leveraged real estate

They would have been better off diversifying and growing slowly but steadily.
Every day you see people making millions and billions of dollars. The internet is filled with look-at-how-successful-I-am gurus. But how many of these gurus stay where they are? Life is long, and there are many opportunities to screw up along the way.
In the age of the internet, it’s harder and harder to “win” in life by being smarter than everyone else. Information is freely available, and anyone can learn smart skills very easily. Everyone is trying to be better and better (e.g. taking online courses). But at the same time, it’s easier and easier to win in life by being less stupid than everyone else. (Just think about how more and more people are getting university degrees while at the same time, more and more people are wasting their time on social media or keeping up with the Joneses).
The average wealthy individual has most of his wealth in his own business (this makes sense, since entrepreneurs are generally the wealthiest individuals). They keep plowing their profits back into the business, hoping to grow and grow at high rates. However, it’s actually better for a business owner to invest most of his profits into the S&P 500. (This is true for business owners as an aggregate). I have seen countless business owners work and reinvest all their profits. But then something changes (the industry shifts or economic forces change), and they are wiped out. No one thinks it can happen to them until it does. This is a hard pill to swallow, especially when someone is on a hot streak. It’s hard to be humble when you are on a streak and say “I’m going to diversify and no longer believe in what I’m doing.”
A business owner’s biggest fear isn’t making less money. His biggest fear is becoming irrelevant as innovations and industries change. No one thinks they will be made irrelevant one day until a ball hits them from left field. Just ask GE and IBM, 2 giants that in their heyday nobody thought could fail. Today, IBM is bleeding slowly and GE is on the verge of collapse.
By investing in the broad stock market, business owners guarantee that they will never be made irrelevant. They can create a company that never dies, like Warren Buffett’s Berkshire Hathaway.
History
The book explains how financial experts have sold market timing strategies forever. There is nothing new about 21st century internet gurus and their subscription services. The book explains that most of these strategies end up not doing too well.
From 1949-1950, there were popular strategies that used various methods to time the stock market’s cycles. These strategies were popular because they sounded logical (sell on the way up, buy on the way down) and demonstrated excellent backtested results.
Unfortunately, most of these strategies performed poorly in the 1950s because the stock market surged throughout that bull market. Those who sold were left sitting on the sidelines.
Psychology
Ben Graham states a point that Warren Buffett reiterates throughout his life

To be a successful investor, you cannot be greedy or envious of others

Judging yourself by how well a bunch of strangers are doing will lead to disaster. E.g. if you watch strangers make a killing buying internet stocks from 1996-1999, become jealous, and think “I should be making a killing as well”, acting on your jealousy would be disastrous in the ensuing dot-com bust.
Psychology
I have said this many times:
Do not “eyeball” a chart. Traders love their charts. But the human eye tends to see what it wants to see when staring at a chart.
The book explains that humans are pattern-seeking animals. We are hardwired to see patterns, even when these patterns don’t actually exist.
Beware of “loss aversion”. The fear of losing $ is twice as great as the greed from making $. This is also why you should be careful with the media. The media knows that bad news is much more click-worthy than good news. That’s why the media constantly blows things out of proportion and exxagerates bad news. Compare the 2 headlines.

  1. The Dow PLUNGED 300 points today
  2. The Dow fell -1% today

These 2 things are the same thing, but the sensationalized headlines are much more likely to attract your attention.
Reading financial reports
The book provides one main suggestion when reading a company’s financial reports:
Read backwards.
Anything that a company doesn’t want you to find is buried in the back of a financial report, which is exactly why you should start there.
How many mutual funds outperform
Financial professionals like to talk about how smart they are. But how many of them actually outperform buy and hold?

*Notice that the longer the time frame, the more professionals underperform buy and hold.
This is not to mock the professionals, but to keep us humble when thinking that we can beat the markets.

Latest data

Factor investing has been all the rage over the past 15 years. Countless financial experts have made millions and billions by convincing investors that investing in factors = outperformance, and then selling them factor-based financial products such as ETFs.
Once again, this proves that in the finance industry, the easiest way to make money is to say things that sound smart, popularize it, and then sell the hell out of it.
Research Affiliates (one of these factor-investing firms) recently came out with a very honest paper examining how well these factors work in real time. (It’s not often you hear a guru or salesman say “I might be wrong”).
For those who are not aware, extensive backtesting has shown that investors can outperform the S&P 500 by tilting their portfolios towards certain factors. We discussed this herehere and here
All of these factors are “proven” to outperform the S&P through extensive historical backtesting.
The paper’s main point is that not a single “proven” factor has outperformed in the past 15+ years. All of the “outperformance” occurred before these factors were researched and popularized.
(Bloomberg has a good article on how quants are fighting for their lives).
The paper explains that this lack of outperformance comes from a few things:

  1. Some of these factors are data-mined. E.g. there is no sound, logical reason for why these factors will outperform. Data-mined factors and strategies break down with the passage of time.
  2. Many of the best known factors are now crowded, which arbitrages away the outperformance of these factors. The backtested results do not reflect the price impact of investors pouring money into these strategies. When a factor/strategy becomes popular, too many people make the same trades and the mispricing disappears. 
  3. Trading costs are quite high for some factors. Backtesting does not take into consideration trading costs. (E.g. “small cap” stocks are generally more expensive to trade due to wider bid-ask spreads)
  4. Some factors are just repeats of other factors. E.g. “value” and “quality” stocks tend to overlap.

A big problem with factor investing is that you are not diversified. E.g. if you tilt your portfolio towards the “size” factor by buying small cap stocks, your portfolio is no longer diversified. Even if you expose your portfolio towards several factors, your portfolio is still not diversified because you are not exposed to every kind of stock in the stock market.
The paper looks at 48 widely-cited factors. The top 15 most popular factors had a median discovery year of 1990 (28 years of out-of-sample data). The other 33 factors had a median discovery year of 2000 (18 years of out-of-sample data).

Here’s what happens to factors before and after they are published (and popularized)

As you can see, a factor’s outperformance decreases almost immediately after it is popularized.
More importantly, from 2003 – present no factor has outperformed buy and hold when including transaction costs. (Certain factors like the “value” factor performed well in the 2000-2002 equities bear market).
Here’s a closer look at these factors. As you can see, the top 6 popular factors have not outperformed at all over the past 15 years.

Here’s a closer look at the same chart, but from 2003- present only

It’s amazing how prescient Ben Graham was in The Intelligent Investor. His latest edition was written in the early 1970s, when he said “strategies that seem to work well in the past will stop working once they are popularized”. It’s like tennis. If you have a specific kind of serve that works well, you will win. But if others start to copy your serve, then your edge is gone.

My thoughts

Here are my thoughts on Ben Graham’s 2 main points:

  1. A lot of strategies that seem “proven” in the past are actually data mined.
  2. If a strategy that outperformed in the past is truly sound, then it will be disseminated and copied very quickly by many other traders. By becoming popular, its ability to outperform will shrink over time.

These points are sobering. I find that reading in general makesyou question many of your fundamental beliefs. The more you read, the more you realize how little you know.
I entirely agree with the first point. Many strategies are data mined. I think all technical analysis strategies (price-only) are data mined.
Here’s a simple example.
The Simple Trading Model works well for the S&P 500. But when you apply the Simple Trading Model to other indices besides the S&P 500, its performance deteriorates.
Every price-only strategy (discretionary or quantitative) works well under a certain type of price action. A strategy that works well in XYZ market will only work well because the most prevalent price action in that market suits the strategy. But we know that the market’s price action changes over the decades. This makes strategies that worked in the past less useful. (And randomly “guessing” the future is even worse).
Here’s a very simple example described in our Technical Indicators Backtest.
This is what happens when you use the simple trend following strategy “buy and hold the S&P 500 when it is above its 2 day moving average”.

As you can see, this strategy performed phenomenally well until 1999. Anyone who looked at this price-only strategy would think “this is absolutely amazing”. But then, out of the blue, the strategy stopped working because the market’s price action changed. A 2 dma strategy works well when the market’s short term has clear uptrends and downtrends. It works poorly when the market’s short term is choppy.
Sometimes price action changes temporarily and soon reverts to its old ways. But sometimes price action changes permanently (e.g 2 dma strategy from 2000-present). You don’t know for certain if a change is permanent or temporary until many years have gone by, after it’s too late. Similarly, you don’t know if a price-only strategy that’s underperforming right now will permanently undperform in the future, or if the underperformance is just temporary.
I somewhat agree with Ben Graham’s second point. Many truly sound strategies that “outperformed” in the past have stopped outperforming because many others replicated them.
But in order for a sound strategy to stop working, it must be widely replicated. So if we find sound strategies that are unpopular among the finance community, then its outperformance should continue.
Nowadays institutional and professional money controls most of the money in the stock market. Hence, retail money is no longer the “dumb money”. The professionals are now the “dumb money” (just look at the average hedge fund’s performance over the past 10 years). Here’s a chart from The Economist

Charlie Bilello conducted an interesting survey asking traders what their favorite market indicator is. As expected, the favorite indicators are valuations, trend (technical indicators) and sentiment (also technical indicators)

This means that if you want to find a strategy that outperformed in the past and will continue to outperform, it must be unpopular. What’s unpopular? Macro. Not many people look at macro in a systematic and quantitative way like we do.
Macro is unpopular because it’s hard to understand. There are thousands of indicators, and many macro indicators often conflict with eachother. It’s not as simple as 2 lines on a chart. Learning how to use macro properly takes time.
Due to the unpopularity of macro, I believe that fundamental+technical strategies will outperform in the future, unless they gain widespread popularity.
Moreover, almost all of our strategies are medium-long term. It’s easier for short term strategies that worked in the past to stop working than it is for medium-long term strategies that worked in the past to stop working. The masses of traders can more easily arbitrage away short term outperformance because it only takes a little bit of $ to change the market’s short term direction. But because the long term is massive (hundreds of billions of dollars hardly have a dent on the market’s long term direction), the long term price-vs-fundamentals relationship is much harder to change.
Above all, this book really makes you consider your place in history. Are we exactly the same as someone in 1940, who thought that they stumbled across “the secret to trading” and then later on found out that it was no better than buy and hold?
I don’t think so, but only time will tell with 100% certainty.
Stay humble.

6 comments add yours

  1. Troy, there is a LOT of wisdom here – and very timely, in my (humble?) opinion.
    Although I trade some strategies that have been very successful over the past few years (mostly short-medium term), I have been increasingly concerned by many of the ideas you cover here.
    When I joined an investment house in the 1970s, who were known for their free-wheeling ways, one of the aphorisms I was told was “speculate to accumulate; once you have accumulated, INVEST!”. It was implied that the former was short-term, whilst the latter was long-term. Overlaying that on some of your points, one could say that a problem most people who indulge that mindset have, is that they stay too long in the first camp, being seduced by “their” success (rather than admitting it may have been only luck).
    Great article – thank you.

    • Very true.
      I have been thinking about the optimal timing. I.e. for a trader, how do you know when to stop speculating and start investing? Or for an entrepreneur like that family friend, how do you know when to stop pushing so hard and start playing more defensive?
      The problem is that it’s very hard to “go out on top”. When things are good, we tend to keep extrapolating the recent past into the future. But on the otherhand, seeing so many others keep doing well from speculating or pushing hard makes us want to continue doing what we’ve been doing.

  2. Very good read Troy. I have been studing the markets sence 1991 and have lived through everything you just wrote. Almost made it big three times, it’s hard to sit on your hands. We all want to be active and use our own brains thinking we can beat the markets. Keep up the good work Troy – thanks –

  3. Nice summary Troy, thank you! In my opinion, all systems (Fundamental and Technical) that backtest have to some degree curve-fit. I think there is no alternative however since our systems have to be developed on real-world data. I suspect trading some forms of trend-following actually become more effective over time if the trend-followers are buying on a similar run-up, reinforcing the rise, then kind of sell as a group, reinforcing the down-trend. Their combined strategy positively reinforces an upsurge in prices and then the fall, creating “waves” in the market. So I think this strategy can become more effective as the word gets out & more people use.

  4. Thank you Troy for this great insight.
    I also agree that AT models are not very reliable, with ONE remarkable exception.
    I would like to talk to you about this, and I have some additional thoughts about your trading models that I will share with you by email.

Leave a Comment