I'm hearing too many resolute, introductory clauses lately:
"Stocks have to..."
"Stocks will definitely..."
"There's no way that..."
This type of thinking can lead to portfolio-tinkering from people who already know they shouldn't tinker; trying to score a win because it feels easy to know what happens next.
It's impossible to know what happens next.
One axiom of evidence-based advisors – those of us who encourage people to invest systematically, without tinkering based on whims, hunches, or opinions – is that during these bouts of drunken volatility, we remind clients the same thing over and over:
At tempting as it may be, DON'T time the market. It won't work.
But you know what? It's a lie.
Timing the market will most likely work.
Just don't do it. Here's why...
The Martingale Betting System
Pursuing most likely, positive outcomes is a terrible investment strategy if the impact of less likely, negative outcomes, is catastrophic. And that's exactly the type of potential negative outcome that market-timing delivers.
It's a well known risk/reward profile.
John Henry Martindale was an 18th century London casino owner. He liked to mingle with his customers, and would encourage them to double-down after losing. He knew how appealing it was for a bettor to quickly recoup losses.
Imagine we are in Las Vegas at a blackjack table. Minimum bet is $5.
You're willing to lose $1,000. What are the chances that you'll be up money at some point?
Played properly, Blackjack odds are about 42% that you win, 8% you tie, and 50% you lose. For simplicity, let's just assume a tie is a loss. Even then, it's a reasonable outcome that you will win sometimes (42%).
Let's assume the goal is to eventually just be in the money...here's how you'd bet every hand so that you could try to be net positive – the chart assumes you lose each hand, but if you'd won, then your net winnings would be the far right column.
Just keep doubling-down.
By the eighth bet, assuming you lost every one, you've now lost $767 and there is no longer enough money out of your $1,000 limit to become positive in one more bet. So we will end there, and identify that as "going broke" (even though you still have have $233 left over that you weren't able to double-down with).
But if you win any of the first eight bets, each which has a 42% chance, you are in the money (though, per the right column above...not by much).
The conditional probabilities of just trying to be in the money by making multiple 42% chance bets, and doubling-down after a loss, are in the middle column below:
You can walk into a casino with $1,000, place a $5 bet, keep doubling-down if you lose, and have a 98.7% chance you will, at some point, make money.
So why don't more people do this?
Because magnitude matters as much as likelihood.
It's easy to win, but you won't win much. It's hard to lose, but you lose everything.
Through the years, the term Martindale has evolved into just being called martingale, and amongst gambling circles – it's a well documented, centuries-old strategy. Just keep doubling-down.
"The martingale is as elusive as the soul."
Alexander Dumas, 1849
Photo: Culture Club/Getty Images
It's critical to restate what winning means here, because it's not good fortune nor vast wealth. It's just being in the money by a small amount. In our example, it's + $5 if you win on hand #1, or + $1 if you win any of hands #2 through #8.
Have we engineered a way to reverse the casino house's odds?
Could we use this system to easily get rich?
Yet we can play so that there is a 98.7% chance of beating the casino?
There's a glaring issue if we want to get rich using a martingale strategy.
98.7% is not 100%.
We haven't altered the odds of each individual game, but simply risked more on subsequent bets to try and quickly recoup losses.
There is a tradeoff.
You cannot increase your odds of success without making commensurate increases to the size of your potential loss.
Usually, if you want to gamble $1,000 by playing a single blackjack hand, you have a slightly less than 50% chance that you'll double your money, and a slightly above 50% chance that you'll lose all your money.
But with a martingale strategy – if you want to gamble $1,000 by playing potentially eight hands each round, doubling-down each time until you win (and then starting a new round), you have a 98.7% chance every round of making $1 or $5, and a 1.3% chance of losing $1,000.
But the 1.3% is an absolute landmine.
And so if you only want to potentially go broke by also having a chance to get rich, you'll have to collect a lot of $1 and $5 wins. That's a lot of playing, and a lot of 1.3%-chance-of-losing-everything opportunities.
So far I've just showed that if you start with a $1,000, you can have a high probability of walking out of a casino in the money.
You can find similar risk/reward profiles in investing.
You can probably time the market.
Let's define "timing the market" as a decision to sell stocks, and intending to get back in at a lower price.
Similar to the blackjack table, let's call winning just being in the money, i.e. once you sold stocks, you had a chance to buy them back at a lower price.
For simplicity, let's use the S&P 500 as the market.
Here's why you can probably time the market:
Since 2012, there were nine years where if you had sold on the first trading of the year, you would find yourself in the money at some point that same year, with the opportunity to buy back in at lower prices.
Only two years, 2012 and 2021, experienced an outcome where the lowest price of the year was the first trading day, and the market just went up after that; i.e. if you'd hoped to ever get in at a lower price, it never happened.
For 2021, though, if you'd just waited until 2022 – as we're experiencing right now, you would have had the opportunity to buy in at lower prices.
It's a pretty helpful data point to think about market-timing:
Since 2012, there is only one "first trading day of the year" where you didn't get another day in the future to buy at lower prices.
And you can imagine this example akin to a martingale strategy in blackjack. Once you're finally in the money, you lock the win and start the process over.
For investing, it could look like this:
Sell stocks on the first trading day of the year
Re-enter stocks at lower prices, whenever they present
Repeat on the first-trading day of the next year
It makes sense that lower prices typically come. Whether stocks go up or down on any given day is essentially a coin flip (about 52% up-days). And because stocks are volatile, they frequently zig-zag, which can be accommodating to market-timers.
If you look at the chart above, you'll see big slugs of time where if you'd exited (like late 2017), you found a lower price many years down the road (like 2020).
That's nothing compared to decades before: if you sold the S&P 500 on the first trading day of the year in 1998, you could have entered at a lower price in 2009, 12 years later:
But not just any point in 2009.
And this is where the risk profile of low-probability but devastating negative outcomes comes into play...the same risk profile as the martingale strategy.
There were 2,830 trading days from the beginning of 1998 until March 9, 2009.
Everyday single day, for twelve years straight, you had a lower entry point ahead.
And then one day you just don't. March 9, 2009. Stocks never looked back.
And such is market-timing and the martingale. So many opportunities to be right and make a little money, and a handful of times that will unforgivingly gut you.
Don't avoid market-timing because it won't work. It will probably work at some point.
Avoid it because you don't need this tradeoff.
We never know when stocks will never look back again.
Likelihood: 2,830 trading days in a row it worked. Then one day it didn't.
Magnitude: Never came back.
Don't do it.