Stock market reporting is broken

Actually, it's not just the reporting. Recommendations are broken too. Deeply. Fundamentally. It is a structural failure of the highest order.

In ninth grade civics class we we did a unit on the stock market. They taught us how to read the daily stock quotes in the newspaper (this was before the internet. Yes, I'm that old.) And our homework assignment was to take $10,000 in pretend money and put together a stock portfolio that we would track for a week. At the end of the week, whoever made the most "money" got some sort of a prize (I don't remember what it was. Yes, I'm that old.)

The kid who won got a 200% return in one week. He did it by "buying" a penny stock at 1/8 (they still used fractions in those days. Yes, I'm yada yada yada...) and "selling" it two days later at 3/8. It seemed like magic. If you could just figure out how to "pick the winners" you could make money. Lots and lots of money. Money money money money money. And so I, like so many before and after me, started looking at stock charts and thinking wistfully about how to obtain a copy of tomorrow's newspaper.

What they didn't tell us, and what I didn't figure out until many years and many lost dollars later, was that the stock market is (imagine this) a market, that is, a place where people come together to buy and sell stuff, in this case stocks. There is no magic. It's fundamentally no different than any other market. At root, the NYSE is just a bunch of folks with lemonade stands, except instead of lemonade they are selling stocks.

So what is it about the stock market that gets people so much more excited about it than lemonade stands? After all, you can make real money selling lemonade (or at least a synthetic brew that vaguely resembles lemonade, but that's another post). Two things. First, stocks are a virtual good, so they are mechanically and logistically easier to deal with than physical goods like lemons and sugar. You don't have to worry about transportation or storage. All you have to do is buy and sell. And this leads to the second attractive feature: price volatility.

Wait, what? Isn't price volatility a bad thing? Aren't we all getting nauseous from the market's recent roller coaster rides? Well, you might be getting nauseous, but I guarantee you that there are some traders out there -- the ones who have the balls or enough inside information to buy low and sell high -- who are making tons of money. And that is exactly what attracts people, the possibility of getting rich quick if you can just guess better than most people which way the market is going to move. Because of various logical fallacies that humans are chronically prone to (like confirmation bias) it is easy to convince yourself that you can do this. But by definition most people can't.

This fundamental irrationality is insidious. It has given rise to an entire industry carefully designed to take advantage of it. OK, maybe you aren't better than the next guy at predicting the market, but surely a professional who has studied the math can be, so if you can't call the market you can do the next best thing and hire someone who can. Vast numbers of people who would never dream of playing the market on their own hire self-styled "financial advisors" to do it for them. And it seems plausible that such talent should be available for hire. But the evidence is overwhelming that it isn't. Study after study has consistently shown that professional money managers cannot consistently beat the market, not even a little bit.

Here's why. Take a look at the two charts at the bottom of this page. (Go ahead and click on the link. It should open in a new window.) This site is a market for bitcoin but that's irrelevant for this discussion. What matters is the content of the graphs. The first one is a history of recent trades. This is the kind of chart that anyone who has ever followed the market is familiar with. It has the familiar ups and downs and twists and turns that make it oh so tempting to try to discern a lucrative pattern that everyone else has missed.

But the second chart is one that you probably haven't seen. It's a market depth chart. It shows a snapshot of the current order book for bitcoin on this exchange. It shows you how many bitcoins people are currently willing to buy and sell at various prices. The green bids on the left are the orders to buy, and the yellow asks on the right are the orders to sell. Notice that the order volume on both the buy and sell side taper off and fall to zero at the exact same price that the last trade of the day took place at (look at the extreme right hand side of the "recent trades" chart.) This is, of course, not a coincidence. If the two sides of the order book had any overlap, there would immediately be trades consummated that would eliminate it.

Now, it is important to note that the market depth chart is not a time history like the recent trades chart is. It is a snapshot of the state of the market at a particular instant in time. The actual history of the market is not the history of consummated transactions, but the history of the order book, which is vastly more information than the history of consummated trades. The American stock market generates terabytes of history data every single day.

And behind the order book there is an even bigger reservoir of hidden state, which is all of the market analysts and individual traders making decisions about whether to buy or sell.

(At this point stop an ask yourself: what would have happened if my ninth grade colleague had done his trading with real money.)

We are now in a position to understand why reporting on the stock market is so utterly broken. Let's start with the evening news. It's probably safe to say that there is not a single general news outlet that does not dutifully report the closing price of the Dow Jones Industrial Average every single day, and the absolute difference between that price and the previous day's closing price. Sometimes they'll leave out the price and only report on the difference. Even on NPR, which really ought to know better, you will hear regularly throughout the day, "The Dow is up by 27 points, the NASDAQ is down by 3." On tonight's evening news the 420 point drop in the Dow will almost certainly be the lead story, with commentators swooning over the fact that this is the ninth biggest point drop ever in the Dow's history or whatever it is.

All of this is almost completely bogus. For starters, the Dow is only a sampling (albeit a fairly representative one) of the overall market. For afters, the closing price is just the price at which the last trade of the day happened to take place. It is a single data point from a large collection of data points (the trade history) which in turn is driven by an even larger collection of data points (the order book and its history) which in turn is driven by an even larger collection of data points, many (most) of which are simply inaccessible.

The bogosity doesn't stop there. Tonight you will hear that so-and-so-many trillion dollars of wealth were "wiped out" by today's Dow drop or some such nonsense. No, they weren't. The value of the market is NOT the closing price multiplied by the number of shares. That assumes that you could sell all the shares for the closing price. If you did that experiment you would find very quickly that this assumption is false. Stock prices obey the laws of supply and demand just like everything else, and if you flood the market with product the price will inevitably go down.

If you wanted to mislead people about what really goes on in the stock market you could hardly do better than to focus their attention on the current state of the Dow relative to yesterday's closing price. And yet that is what every single news outlet does.

Why?

Well, part of it is simple inertia. Daily reporting of stock market closing prices goes back decades. I can remember watching the CBS Evening News with Walter Cronkite (when I was a kid -- I'm not that old ;-) and even Grandpa Walter was doing it, except that back then the mantra was, "The average price per share on the New York Stock Exchange gained an eighth of a point, and a quarter point on the American exchange." Today so many people pay attention to the Dow that it actually matters because of the psychological effects that big Dow swings have on people, which in turn can create market movement on its own. But the closing price of the Dow doesn't really matter. The Dow by now is kind of like Paris Hilton, famous only for being famous. If ever there was a tail that wagged the dog, the daily closing price of INDU is it.

But part if it is because there are people who make money off of your ignorance. Vast, vast piles of money. The "financial services industry" (I put it in scare quotes because it's quite possibly a net drain on the economy at this point) is about 20% of the U.S. economy. Not all of that is waste. You can't have capitalism without capital markets. But around the core capital markets that enable a modern industrial economy has grown a vast parasitic ecosystem of "financial advisors" whose livelihood depends on convincing you that they know more about the markets than you do.

And in fact they do know more about the markets than you do. They know that the best way to make money in the markets is to convince people that they know how to make money in the markets (whether or not it's actually true) so that they can play the market with your money instead of theirs. And collect fees from you.

To be fair, there is one -- and only one -- market strategy that does work. It's called asset allocation, and it can give you a small edge in terms of risk-reward over the market as a whole. The math behind this strategy is a little hairy, but not beyond the capabilities of any reasonably bright math undergrad. But anyone who tells you that they can do better than that is a flim-flam artist. (And anyone who gets a hundred million dollars a year to do this work is a crook.)

So all of those stock picks? Buy and sell recommendations? Almost entirely bogus. And you can tell that they're bogus by the form that they take: buy or sell. A recommendation to buy or sell is meaningless out of context. The only kind of recommendation that could possibly make sense is a recommendation to buy or sell within some particular time window, below (for buying) or above (for selling) some particular price. Every stock is worth buying at some price (unless the company is actually bankrupt). And every stock is worth selling at some other price. A stock recommendation that even had a chance of being worthwhile would take the form of "We believe the actual market value of this stock over a time interval T is between X and Y. So if during T the price is below X, buy. If it's above Y, sell. And if it's between X and Y hold." Or something like that. Whether or not you should actually buy or sell depends on a ton of other factors, like your tax situation, whether or not you need cash, your tolerance for risk, etc. etc. etc. But a naked buy/sell recommendation cannot possibly have any content, and study after study shows that this is so.