Chapter 11: Stop-Losses Stop Losses
“A stop-loss can stop your losses!”
EVEN THE name stop-loss sounds good. Who doesn’t want to stop losses? Except, stop-losses don’t do what people want them to. Instead, they tend to trigger taxable events and more transaction fees. And they more often stop gains than they do losses—on average over the long-term, they’re a money loser. Don’t buy into this costly myth.
Stop-losses come and go in popularity. You tend not to hear much about them in the latter stages of a bull market. Sir John Templeton famously said, “Bull markets are born in pessimism, grow on skepticism, mature on optimism and die of euphoria.” Stop-losses are mostly a pessimism–skepticism game, though they have adherents no matter the market cycle. They tend to appeal to people who think downside volatility is bad but upside volatility isn’t volatility at all. But as discussed in Chapter 4, you can’t get the upside without the downside.
The Stop-Loss Mechanics
For the uninitiated, a stop-loss is some mechanical methodology, like an order placed with a broker, to automatically sell a stock (or bond, exchange-traded fund [ETF], mutual fund, the whole market, whatever) when it falls a certain amount.
That amount is up to you! There’s no “right” amount for a stop loss (mostly because there’s no level proven to improve long-term performance). Typically, folks tend to pick round numbers like 10% or 15% or 20% lower than their purchase price. No reason— people just like round numbers. Nice and neat. They could do 11.385% or 19.4562%, but they don’t. No statistical reason why 20% is better than 19.4562%.
The idea is the stop-loss is supposed to protect investors from big downside. If a stock drops and hits the stop-loss level, it’s sold. No big 80% drop disasters. Which sounds appealing! Who doesn’t want to stop losses?
But they just don’t work—not like people hope they do.
If they worked, every professional would use them. If they netted better gains with limited downside, that’s a money manager’s dream. It would make clients more money. More money for the client is more money for the manager. Win-win-win! Yet, I’m not aware of any major longtime successful money manager who uses them—not even occasionally. I have no doubt some financial salespeople may promote them. Not because they improve performance (because provably, they don’t). But stop-losses force sales, and if you’re paid by the transaction, using stop-losses is one good way to increase the number of transactions. Good for the salesperson, but a conflict of interest and not optimal for the client.
Stock Prices Aren’t Serially Correlated
To believe stop-losses work, you must believe stocks are serially correlated. If something is serially correlated, it means past price movements predict future price movements, i.e., a falling stock will keep falling, and a rising stock will keep rising.
There’s a school of investing built around this idea, called momentum investing. Contrary to a vast body of scholarly research (not to mention actual empirical evidence), these folks believe price movement is predictive. They buy winners and cut losers. They look for patterns in charts. But momentum investors don’t do better on average than any other school of investors. In fact, they mostly do worse. Can you name five legendary momentum investors? I can’t think of one.
Stop-losses and momentum investing don’t work because stocks aren’t serially correlated. A price’s movement yesterday on its own has zero impact on what happens today or tomorrow.
Stocks that fall a certain amount—whether 5%, 7%, 10%, 15%, 19.4562%—aren’t more likely to keep falling. Yet stop-lossians act as if that’s so. Think it through: Would you only buy stocks that had risen a bunch? Instinctively, you know that wouldn’t work. Sometimes a stock that rises a lot keeps rising, sometimes it goes down and sometimes it bounces along sideways. My guess is most people get this in their bones: What goes up doesn’t necessarily keep going up. So why don’t folks understand that correctly on the downside?
Certainly, stop-losses appeal to that caveman part of our brains that hates losses more intensely than it likes gains (what’s known as myopic loss aversion). But falling prey to evolutionary responses hurts much more often than it helps in investing. Who wants to invest like a caveman?
Pick a Level, Any Level
Suppose you wanted to do stop-losses anyway, against my recommendation and contrary to the industry standard investing disclosure that “past performance is not indicative of future results.” What level would you pick? And why? Suppose you picked 20%, just because you like the number 20. (It’s as good a reason as any other reason to pick a stop-loss level.) When a stock drops beyond that amount, it triggers your stop-loss, and you automatically sell.
But it’s basically a 50–50 chance it continues dropping or reverses course. You’re trading on a coin flip. Coin flips make bad investment advisers.
For example, the stop-loss won’t stop you and say, “Hey, why do you think that stock dropped 20%? Was it because the entire market corrected that much, and your stock went along for the ride?” Market corrections are common—happen about once every year. If a stock drops with the broader market, that’s not necessarily the stock’s fault. The stop-loss then doesn’t protect you against loss; it just guarantees you sold at a relative low and paid another transaction fee. You might be sitting in cash when the market—and your now-sold stock— reverse course fast and zoom to fresh highs. This is buying high, selling low.
And what do you buy next? Maybe the next stock you buy also eventually drops 20%, triggering another stop-loss. Repeat. Repeat. You can buy 20% losers (or 10% or 19.4562%) all the way to zero. Just because you use a stop-loss doesn’t guarantee your next purchase will only rise. And maybe the initial stock you automatically sold reversed course and zoomed up over 80% in the next year. You missed that! You sold at a relative low, paid two transaction fees and missed out on the good part. Maybe you tell yourself you’ll buy back once you think the trouble has passed, but I say, “Nonsense.” If you sold automatically, what fundamentals are you looking at to tell you to buy back in? And if you could somehow know with certainty when trouble had passed, you wouldn’t need a stop-loss at all. In fact, you’d probably already be unimaginably rich and wouldn’t need this book.
Here’s another way to see this. Suppose you buy XYZ stock at $50 and it zooms to $100. Then your friend Bob buys it, and it drops to $80—down 20%. Should you both sell? Or just him with his higher cost basis? For him the stock is down 20%, but for you, it’s still up 60%. Does that mean the stock is ok for you, but not ok for him? Why?
That’s the problem with stop-losses. There’s no answer to “Why?” other than, “Just because.” Just because isn’t a strategy.
Stop-losses don’t guarantee protection against losses. They do increase the odds you miss out on upside, and they definitely increase transaction costs—perhaps why some brokers have never stopped promoting them. There’s no evidence they produce better results, but there’s mountains of evidence to the contrary. Better to think of them by the name that describes them better: stop-gains. Stop yourself before using stop-losses.
This excerpt from Little Book of Market Myths is distributed by permission of Wiley & Sons and is subject to copyright laws and other restrictions imposed by the provider of the information.
The books and articles mentioned on this site reflect opinions, viewpoints, analyses and other information of the author, which are subject to change at any time without notice. Excerpts are provided for illustrative and informational purposes only. Nothing in these books or articles constitutes investment advice, or any recommendation that any particular security, portfolio of securities, transactions or investment strategy is suitable for any specific person. Information contained in these books or articles should not be regarded as a description of advisory services provided by Fisher Investments. Investments in securities involve a risk of loss. Past performance is never a guarantee of future returns. Not all predictions were, nor future predictions may be, as accurate as those presented herein.