What is priced in to stocks and other securities?
Most people today have come to believe that markets are "efficient" ... meaning that prices reflect publicly available information. We do not believe this. No active investor does. But every active investor should have a theory for what is priced in...and that is the topic of this article.
Ultimately, the goal of active investing is to spot value. That is, what securities are worth more or less then their price. We call this a "mis-pricing" and investors use their measures of it in hopes of capturing "alpha"...the amount by which they outperform their benchmark.
The debate over how to identify value will never end. Why?
Because no one can ever prove that outperformance was due to a true convergence of price to value...or just luck from unpredictable events like a pandemic, new CEO, act of GOD...whatever. The point is we never see the true "intrinsic value" ... all we can see is the price, and all the guesses as to what people think the price should be.
This is why the debate over efficient markets can never be resolved. It is, however, necessary for an investor to have a view on market efficiency because this view is a first principal that will shape investment strategy.
The Efficient Market Hypothesis
The "Efficient Market Hypothesis" has been around a long time, but really picked up steam over the past two decades. Jack Bogle at Vanguard led the way as investors pilled into low cost index funds. These funds provided diversification with only one catch...you were guaranteed not to outperform the index. On the plus side you would not underperform either. The trade-off gave investors comfort that they need not know anything about what they were investing in because hedge funds and other institutions already removed all the inefficiencies in markets. No point in doing any research if you can't beat the experts. Or so goes the narrative.
Benjamin Graham, father of "Value investing" in his own words explaining why before his death he came around to the idea that markets are mostly efficient.
“I am no longer an advocate of elaborate techniques of security analysis in order to find superior value opportunities. This was a rewarding activity, say, 40 years ago, when our textbook "Graham and Dodd" was first published; but the situation has changed a great deal since then. In the old days any well-trained security analyst could do a good professional job of selecting undervalued issues through detailed studies; but in the light of the enormous amount of research now being carried on, I doubt whether in most cases such extensive efforts will generate sufficiently superior selections to justify their cost. To that very limited extent I'm on the side of the "efficient market" school of thought now generally accepted by the professors.”
Those are strong words coming from the man that inspired Warren Buffett to become the most widely respected value investor living today. Of course, Warren hasn't beaten the market over the past decade and is doing quite poorly so far in 2020...the recession year that many investors in Berkshire had waited a decade for just to underperform the S&P 500.
So perhaps are markets efficient?
To answer this we first need to understand what this really means...
There are several efficient market hypothesis (EMH). All of them essentially try to explain what information is in the price and what is not. Two forms are particularly well known.
Weak form (W-EMH) presumes that historical prices are incorporated into price. If true, investors should never bother looking at price trends and focus entirely on examining fundamentals such as cash flows and credit ratings. Questions like, "Is the price down a lot?" are pointless according to W-EMH because believers say that if the price is down...its down for a good reason.
Strong form (S-EMH) presumes that historical prices and fundamentals are incorporated into price. If true, investors should never bother examining any information at all. S-EMH is the culprit behind today’s conventional wisdom that reasonable investors not bother thinking too deeply about their investments. Questions like, "Is the PE ratio really high?" are pointless according to S-EMH because believers say that if the valuation is high...its high for a good reason.
We believe that markets are “Algo Efficient”(A-EMH). What this means is that easily
measurable information such as price-to-earnings ratios, revenue growth rates, credit ratings, and profit margins are largely incorporated into price.
Our hypothesis is supported by the fact that unlike in decades past, this information is now easily available via the internet. An increasing portion of capital today is also being allocated directly by machines and the information fed into machine learning algorithms is going to largely consist of what is easily measurable.
Can we prove that markets are A-EMH?
The reason is that we are essentially stating that easily quantifiable data is in the price. This leaves out all the information that is messy and hard to pinpoint, aggregate, and include in our own algorithms. Thousands of quants poor over data trying to beat markets, but its really questions like these that, if answered correctly, lead to outperformance...
Is Elon Musk is a fraudster or a genius?
Will households still want to take a cruise again after the pandemic?
Will we all want to keep working from home forever?
Could China challenge the USA as a leading superpower?
Can gasoline powered cars challenge the rise of EVs?
Will solar power be a cheaper for of electricity than natural gas?
Will Democrats and Republican's come together after the election...or for the most part keep treating the other like enemies instead of fellow Americans.
What impact will the election have the willingness of businesses to take risks?
Will consumers care more or less about SARS-CoV-2 over time?
These are tough questions...messy questions...but thoughtful people willing to challenge conventional views stand a decent chance of coming to the right answer.
We can't prove that uncertainty regarding these types of questions are what rives mis-pricing...but so far it seems to be working.
What are the implications of Algo Efficient Markets?
We make the following key assumptions in our investing process...
1. Prices already reflect easily quantifiable information.
2. Buying value requires understanding.
3. Alpha is generated by anticipating the machines.
Easily quantifiable information like past prices, price multiples, cash flows, and macroeconomic data like unemployment rates, GDP, and breakeven inflation rates is already baked into the price of financial assets. Hundreds of hedge funds and other financial institutions employ thousands of statistical models to identify mis-priced securities. Their influence on market prices is enormous. Retail investors only make up about 20% of trading volumes. The rest is run by professionals and they increasingly use algorithms. The space is now crowded...so where anomalies once meant mis-pricing they now mean the model is wrong. That's why hedge funds lose to the S&P 500 nearly every year (19 of the past 20) and why only a handful of investors that focus on understanding rather than statistics are beating markets. Three examples include Stanley Druckenmiller, Ray Dalio and Cathie Wood ... whom we follow closely.
For this reason, investing in companies based solely on the examination of easily measurable information like price trends and P/E ratios is a waste of time. AQR is perfect example of this. AQR has underperformed the S&P 500 for the past two years using strategies that use only easily measurable data and without any consideration for the actual nature of the underlying business. AQR is run by Cliff Asness, understudy of Nobel Prize winner Eugene Fama, and widely regarded as one of the greatest quantitative investment strategists in the world.
Buying value now requires understanding. This was not always the case. Back in Benjamin Graham's day some good data on a company could be enough to justify a big bet. Buying companies with low P/E ratios was a great strategy...until everyone started doing it.
Warren Buffett has underperformed the S&P 500 by a significant margin (13.78% compared to 15.39%) over the past 10 years. No one can prove why, but we think its because of two reasons. First, Buffett tends to avoid investing in the technology sector because of the complexity. He prefers mature and profitable companies that pay dividends in stable businesses. However, these same companies also tend to carry lower risk premiums precisely because they are generally believed to be “safe”. A second reason is that nearly everyone and their grandmother believes in “value investing”. One result of this demand is a plethora of “value” ETFs making it easier for investors to passively bid up the price on companies with high dividend yields and lower price-to-earnings ratios.
But times...they are a changing...
Alpha today can only come from one source: Anticipating the machines. Eventually the machines will catch up to reality. Ever wondered why stocks didn't react to COVID19 until the end of February...a month after news of the pandemic broke out? Ever wonder why stocks waited a month after enhanced unemployment checks got cut at the end of July before correcting? Ever wonder why Tesla's price exploded after it became profitable even though their costs had already been predictably shrinking for years do to the employment of new technology and economies of scale?
Simple...Algos read data...and data is backward looking.
We are employing the Algo Efficient Market Hypothesis right now in our big COVID19 short. On our portfolios page we have listed about 15 companies that we believe have benefited enormously from the pandemic in ways that will not entirely persist. Their share prices are sky high...as any algo taking in the jump in sales would justify. But Algos don't know that the mortality rate is dropping, that 11 vaccines are in phase 3 trials, and that most Americans are wearing masks when they can't socially distance. Algos just read the data as it comes in and adjust accordingly. We bought some cruise lines this week for the same reason.
There are two ways to accomplish getting ahead of algos. The first is the use of our own algorithms for identifying historical patterns in easily measurable information such as historical prices. A good example of this is “Death Cross” a signal of falling momentum that signals the end of a bull market. Even if people stop investing with a herd mentality the algorithms will likely still act on this historical signal and cause future corrections...making the signal self-fulfilling. This happened predictably in late 2018. As soon as the death cross occurred algos hit the sell button. It doesn't always work...but you can calculate the historical returns for yourself if you like. We did.
The other way to do this is identify material aspects of companies and industries that are hard to measure, but likely to influence future prospects for the company. A deep analysis into company moats, falling cost curves, and adoption rates is behind a lot of our investments. That's why we doubled down on Tesla last year. That's why we are so focused on Genomics and 3D Printing right now.
We spend a lot of our time anticipating algorithms. It is arguably the cornerstone of our investing philosophy. We are constantly refining our thesis, but we seem to be on the right track.
Thank you for reading our thesis on Algo Efficient Markets! If you enjoyed this then you may want to follow us on Twitter, Medium, or SeekingAlpha.