Death of price discovery?

Foreword

I first made a version of this post over a year ago in the early days of 2020 BC (Before Covid19). Back then the markets were roaring, despite news of some virus in some province of China. I mean, who cares, right?

Just before news of the virus, we just sort of started a trade truce, for a trade war with China, that wasn’t really a war, officially called a trade dispute. And that itself was just a few short weeks after one of the most tense nuclear standoffs between the US and North Korea in recent memory. Oh, and 2020 was an election year. I mean, clearly, markets should be ROARING! What could go wrong? Amirite?

As usual, a reminder that I am not a financial professional by training — I am a software engineer by training, and by trade. The following is based on my personal understanding, which is gained through self-study and working in finance for a few years.

If you find anything that you feel is incorrect, please feel free to leave a comment, and discuss your thoughts.

Death of price discovery?

Something that has been catching the headlines lately, and was actually talked about in certain circles for a few years now, is the “death of value”.  I want to broaden the scope a bit, and make the claim that (not that I believe it, but rather, just as a motion to debate) “price discovery is dead”.  Not “dead” as in completely irrelevant, but “dead enough” that maybe it does not (and maybe should not!) factor into most people’s investment decisions, and certainly means valuation metrics like P/E, P/EG, P/S, P/FCF, etc. ratios are mostly meaningless nowadays.  Whether it eventually resurrects, is, also, an interesting side topic to explore.

Let’s go back a few decades, to pre-1990.  Especially in the 80’s, large financial entities that engaged in essentially “price discovery” were common.  Large hedge fund do fundamental research (Berkshire Hathaway is a well known example) and buy/sold based on that research.

In the 90’s and beyond, new financial innovations and research popularized “passive investing”, where essentially, the individual benefits from the work of the “price discoverers” via buying a basket of names.  With this as the basis, further research mostly tended to focus on allocation across asset classes, with an almost religious belief that prices properly reflected fundamentals (or at least, close enough to not matter).

Having worked in a (relatively successful) quantitative hedge fund for a few years, it surprised me greatly to find that many there probably cannot make sense of a balance sheet, income or cashflow statement — even people working on the research side of the business don’t always focus on fundamentals.

Also, when talking/listening to people from other hedge funds (some of which are not quantitative) and banks, it does seem like focus on fundamental value is not highly prized.  Most people seem to care more about quantitative issues like “relative value”, “momentum” or “reversion”, technical issues like “crowding” or “short squeezes”, and behavior issues like “story stocks” (aka meme stocks), etc.  In the past decade or so, almost nobody has talked to me discussing a trade where the subject was a buy or sell because of its fundamentals, with an argument that wasn’t easily dismissed as naïve or wrong.

Final note to illustrate this point: I attended a course, over ~6months of roughly fortnightly classes, where a lecturer (former head researcher of a major bank) was invited by a major bank to teach the basics of finance to their new hires, and new hires of clients (i.e: the hedge fund I worked for).  Throughout this course, we learned about how to value derivatives and bonds, how to trade swaps, FX, etc.  Yet, we spent a grand total of 0 minutes on how to value a stock.  Yes, we literally learned how to derive and compute various higher order derivatives of a stock, but never bothered to even discuss how to estimate the price of the underlying.  Imagine if your physics/calculus teacher taught you how to derive dy/dx, d^2y/dx^2, etc., but did not teach you how to solve for y = ax^2 + bx + c.

Quant hedge funds make money by being more efficient, or taking advantage of small arbitrage opportunities of some form (time arb, info arb, statistical based on behavior/technical, etc.), banks profit off spreads based on arbitraging their superior market knowledge or customer relationships.  What about fundamental hedge funds?

Let’s say you have the magic formula, and you can determine the exact correct price for a stock at any point in time (to the 10th decimal, etc.).  What then?  You cannot profit by simply buying/shorting the stock.  For that to work, it’s not enough that you “see the truth“.  Others must see the truth too, and they must see it after you.  This is why many fundamental hedge funds (Muddy Waters is really the only one that comes to mind right now) go public so often with their research. But convincing a large majority of people to see your point of view isn’t easy, and even if you are right, if someone with enough wealth decides you’re wrong, you may still lose financially (see Ackman vs Icahn, re: Herbalife).

So, most non-quant hedge funds nowadays mostly just work off a few related concepts:

• Relative value (1)
• CAPM (1)
• MPT (2)
• Technicals (e.g: selling covered calls to generate additional income, many/most long-short strategies, etc.)

Very few of them actually find stocks that are undervalued/overvalued and then buy/short them.  Underlying all the points above, is the argument that everyone essentially believes that the price of stocks are “correct” (or at least, “correct enough”).

What about the average person?  Here, I have no anecdotal evidence, but general sweeping statements (which I hope most people will agree with):

• The average person knows even less about valuations than the average financial company employee.
• The average person either buys passive instruments and forgets about their investments, OR
• are mostly buying “randomly”, OR
• are mostly buying things they understand (i.e: brands they like).

So, given that very few people are actually trying to value stocks on a fundamental basis, are prices still correct?  But more importantly, do they still matter?

And here, I’m going to make a wild claim (again, not that I fully believe it): Prices no longer matter (much)… for a company that is stable.

Let’s say you have a company, we’ll call it Tongass.  Tongass is not a particularly profitable company, but it is profitable.  It IPO’d a long time ago.  So, how much would you pay for 1 share of Tongass?

If Tongass went bankrupt, then $0 (or more accurately, (total assets – total liabilities) / share count). But since Tongass is profitable, and seems likely to remain that way, bankruptcy is not an issue.

So now what? $1? $2,000? $10^6?

At some extremes ($1 or $10^6), the stock price would matter, since Tongass no doubt rewards some number of employees with equity, and that affects their motivations, etc.  But for the vast range of “in the middle” (say $1,000-$3,000), it probably won’t matter too much.  Yes, some employees will be able to afford nicer houses/cars, but most of them will still need to work, and most of them will still be motivated to remain with Tongass. (3)

So, for a large range of the stock price, the operating metrics of Tongass will not be affected much, if at all.

And if the stock price doesn’t really matter to Tongass, and nobody seems to really care about the absolute value of the stock/company, then, does it really matter?

Footnotes

1. I don’t consider CAPM as fundamental valuation, but rather, a form of “relative value”.  If you look at the formulation, you can see that everything in CAPM deals only in quantitative space, there’s zero mentions of whether the company is even profitable.
2. Similarly, MPT is mostly about trying to manage risk and configuring your portfolio for optimal expected returns based on accepted risk parameters. There are no mentions of fundamentals at all in MPT.
3. Just to note, that if you were actually considering between $1, $1,000, $3,000, or whatever for each share of Tongass, you would have already fell into the trap — I did not mention at any point the metrics of the company, nor how many shares it has outstanding. So how can you even begin to value a single share of an unknown quantity of shares, for an unknown quantity of sales, profits, expenses, etc.?