I have heard many people in my life say things like

The market is low so it's a good time to invest


The market is high so don't invest

It makes me wonder if mean-reversion is actually a thing for long-term passive investing. This essentially implies that these people think they can forecast the future - which is unlikely given what I have read (e.g the famous saying: time in the market instead of timing the market).

If mean-reversion does exist, am I correct to say that timing the market is likely to reap great rewards?

If mean-reversion does not exist:

  1. Why were the returns for the 10 years following the 2008 recession unusually high?

  2. Why do so many people think it's a bad idea to invest when the market is high and a good idea to invest when the market is low?

  • 17
    Investing a large amount of cash into stocks right now probably isn't a good idea. Why?
    – quid
    Dec 3, 2019 at 19:46
  • 9
    Have your acquaintances in your circle who say the market is too high sold their positions in the market? If not, why not?
    – Magua
    Dec 3, 2019 at 21:00
  • 17
    the key thing to note is that "the market is high" is only something you can say in retrospect (so, really, you can only say: "the market was high"). 18 months ago a financial advisor told me that the Dow Jones had to drop before going past 25000, and that the correction was going to be like right now. He is still waiting.
    – njzk2
    Dec 4, 2019 at 5:33
  • 2
    @jamesqf, sure, I'm not debating that there are some plausible rationalizations for maybe not loading up on stock; I'm just debating the blanket statement that "investing a large amount of cash in to stocks right now probably isn't a good idea." I just wanted a justification for that sentence and think that person should have written an answer.
    – quid
    Dec 4, 2019 at 19:09
  • 3
    Try a few rounds of this, to see how hard it is to time the market.
    – Glen Yates
    Dec 4, 2019 at 19:28

12 Answers 12


If mean-reversion does not exist:

It does not.

Why were the returns for the 10 years following the 2008 recession unusually high?

Because after the market finishes going down, it goes up. That's what it means to "finish going down". Similarly, after it finishes going up, it goes down. That's how we know when it finished going up.

So it went down. Then, at some point we could not predict, it finished going down and started going up. Then, again, at a point nobody could predict, it stopped going down and went up.

Why do so many people think it's a bad idea to invest when the market is high and a good idea to invest when the market is low?

It is a bad idea to invest when the market is high and a good idea to invest when the market is low. And when you can see both to the left and to the right on a chart, it's easy to tell when the market is high and when it's low. The problem is that when you can only see left on the chart, you can't tell whether the market is high or low.

Is the market high now? You won't think so if it goes up for the next year. You won't be sure when it was high until it starts really coming down, will you?

  • Mean reversion exists in Pairs trading where the issues are very highly correlated. In fact, the correlation is so high that there is nothing but mean reversion - and the reversion is tradeable both long and short when the sector is volatile. Dec 18, 2019 at 3:55

If mean-reversion does exist, am I correct to say that timing the market is likely to reap great rewards?


Just one problem: you can't time the market.

why do so many people think it's a bad idea to invest when the market is high

Because "they" think that the market will crash Real Soon Now.

and a good idea to invest when the market is low?

Because when it's down, it will eventually go back up.

The bottom line is that unless you #1 pay very close attention to a lot of economic signals, #2 are much smarter than most economists, and #3 have much better than average economic models...

you won't know when High is.

(You can, though, do this with specific stocks, but also takes a lot of time and study.)

  • 4
    To put some data behind this, look here: ofdollarsanddata.com/how-to-invest-a-lump-sum Always buying as soon as possible significantly outperforms techniques such as dollar cost averaging. Waiting for a specific event like a dip is even more likely to lose you money, since dca already outperforms or equals most analysis techniques.
    – niwax
    Dec 4, 2019 at 14:27
  • 2
    "Because when it's down, it will eventually go back up." Part of an issue here is that people don't necessarily have a long term horizon that allows this to be relevant. I'm not saying you can time the market, but you certainly can get burned by getting a short-term investment at the the wrong time. I just find this "it'll eventually go up" justification to be dismissive of many (most?) people's concerns, who have short time scales.
    – Alexander
    Dec 4, 2019 at 15:55
  • 2
    @xyious the Shiller P/E ratio is about 30, and was only 27 in 2007/. (It was 45 in 1999.) Thus, we really don't know when High is.
    – RonJohn
    Dec 4, 2019 at 17:23
  • 1
    @Alexander-ReinstateMonica but many of us have long-term (retirement) visions.
    – RonJohn
    Dec 4, 2019 at 17:24
  • 1
    You might want to revise "you can't time the market", in light of your later paragraphs. Perhaps "Few people can time the market" or something similar.
    – JBentley
    Dec 5, 2019 at 11:22

The common expression in retort would be "Time in the market beats timing the market."

Meaning: On average, the stock market rises [because on average, the global economy is expanding as outputs continue to increase] - a common rule of thumb for North American markets would be 7% / year, after factoring inflation. This general rise beats out the average person's ability to correctly predict exactly when a sharper rise or sharper fall will occur. Doing this correct 'timing', some would argue, comes down to pure luck, or at bare minimum takes a high degree of knowledge if it is possible at all.

Rather, it is better to simply invest early, and often, consistently putting aside money for retirement / other financial goals. The market will ebb and flow, and on average, in the 40+ years you will be working and saving, this will earn you a positive return.

As a quick example of this bearing true: we all know about the 'dot-com bubble burst' in 2000, when the market sharply corrected for significant overvaluations of un-proven tech companies. But the NASDAQ in Jan 1998 on the way 'up' to the top of the bubble was about 1,500, and it never fell that low again until the global financial crisis in 2007 [by April 2009 it was already back up to 2,400, a total gain over that period of 900, or about 100 / year, which equates to about 7% annually if we forget about compounding - and that includes 2 market corrections!]. So if you foresaw the dot-com bubble burst and left the market in '98, you would have lost overall gains even though you correctly predicted the burst!

To paraphrase The Big Short, being too early with a timing call is the same thing as being wrong.

  • 3
    To "win", you can't just leave the market, because money out of the market is money being lost. You have to also get back in. Get it wrong on either side and you can lose a lot of money.
    – Nelson
    Dec 4, 2019 at 2:30
  • 13
    If you left the market in 1998, more likely than not you would be kicking yourself in 2000 and went back in right before the crash. Same as Isaac Newton did in the south sea bubble.
    – jf328
    Dec 4, 2019 at 3:39

No one can predict when the market will correct or how far it will drop but you can certainly react to that if you are not fear based or uninformed. This assumes that it's not a one day event like 1987 (down 22%) when everyone is the proverbial deer in the headlights but rather a 50% drop over 15-18 months (see 2000 and 2008).

If you are concerned about risk, manage it. You can reduce it by diversifying across sectors, countries, growth versus income. You can avoid high beta stocks. You can limit your risk with options, with some protective strategies requiring little to no cost. And for the experienced trader/investor, it's possible to make money during those large corrections, despite what the nayersayers suggest.

Why were the returns for the 10 years following the 2008 recession unusually high?

I'd offer three reasons:

  • What goes down big eventually goes up
  • Quantitative easing
  • Tax cuts

Why do so many people think it's a bad idea to invest when the market is high and a good idea to invest when the market is low?

Why do people like buying things on sale? :->)


Where I live (Finland), stock market is high but not as high as in the US. On the other hand, there is no good alternative for stocks because bond yields are negative! So, compared to the bond market, the stock market is not at all that high.

The problem in general with timing the market is that if you time the market, you spend less time in the market. This, alone, will mean your returns are likely worse.

Instead of exiting the market entirely in situations where you think the market is overvalued, I would recommend another strategy: organize your assets to that when the market is high, the proportion of bonds is higher and the proportion of stocks is lower than in a neutral portfolio. When the market is low, you go all in to stocks.

The neutral portfolio is the portfolio you would choose in the absence of any under/overvaluation indications. How much bonds and how much stocks the neutral portfolio has depends on your ability to take risk, your age, your future income potential, your time horizon for investing etc.

And also, you can use the same trick for geographical diversification. For example, now could be a good time to invest to Finnish stocks, but a worse time to invest in US stocks because of huge overvaluation. So, pick more Finnish stocks and less US stocks into your portfolio.

  • 1
    But there are always other things that can be done with money. For instance, after the '08 crash I basically cut my discretionary spending to the bone so that I could invest. These days I'm spending that money on remodelling the house, and on various recreational things.
    – jamesqf
    Dec 4, 2019 at 6:04

Many stock indexes are capitalized-weighted which makes them momentum funds of current trends. And so the historically recent double-tops or triple-tops don't have a lot of technical meaning. The index will just re-balance on pull-backs and ride the new balance back-up. One potential problem would be that there are no trends but just that everything is down. Now capitalized-weighted indexes are being used by most people as the market average !

Another way of looking at the stock market is P/E ratios. But again that's often the P/E ratios of capitalized-weighted indexes and not the P/E ratio of the market average. Furthermore as profit margins have historically increased then expect that P/E ratios have historically increased.

But a large portfolio of individual stock selections will often have both over-valued stocks and under-valued stocks. Then the investor can historically step in and out of individual stocks of the custom portfolio. There will always be something to buy and something to sell.


Market prices can be high or low compared to the past, but the only thing we know about the future is, that the price is most likely wrong meaning that it's not equal to the true value of the stock which is unknown.

That dot on the chart which indicates the price of a stock is a reflection of the trades that took place at that price. If the price shows $10 for a particular stock that means there was somebody who thought the actual value was lower (which is why he sold it) and somebody who thought the actual value was higher (which is why he bought it). For any "rule" that is supposed to indicate whether one should buy or sell, there has to be someone doing the exact opposite of what the rule says. So why are we using the price to estimate the value of a stock? While we can be sure that it's wrong, it's still the best thing we have!

Timing the market basically means that one has an even better estimate of the value of an asset, which, simple luck set aside can happen because of three possible reasons:

That person is either

  1. smarter than everybody else.
  2. faster than everybody else.
  3. cheating.

Note that this only applies to actually "beating" the market without relying on luck, as you can (and on average will) make money simply going with the market - which is what you should try to do. The best approach to that is diversification. If you buy the whole market, then you will get the performance of the whole market without having to rely on luck for making the best picks. You can also spread your investments over time to reduce risks further.

Slight correction thanks to JBentley:

One last thing: Since nobody can beat the market by timing or picking (again, without relying on luck), it doesn't make sense to pay someone for trying. Paying someone to do the diversification for you is not about performing better but about reducing transaction costs.


It is always possible if you invest a large amount of money now that value will suddenly decrease. However, you don't want to wait until there is a large decrease before investing, because that could be a long time in the future, and depending on what happens in between, the next big drop might even 'bottom out' at a higher level than present day.

The solution is simply to begin investing now, but not all of your liquid savings right away. Invest monthly at an elevated rate, so that after a certain period, perhaps one or two years, you will reach your target liquidity level and then invest at a stable rate from then on.

With this strategy, if there is a 'big drop' soon, you are only partially exposed.

  • 1
    It is debatable whether this technique results in any benefit. Yes, it lowers your risk of buying at the top, but it also increases your risk of buying at the bottom. Meanwhile, in an overall upwards trending market, your money has spent less time being invested than if you did it all in a lump sum at the beginning, leading to an overall loss on average.
    – JBentley
    Dec 5, 2019 at 12:11
  • Very sensible, except that "buying at the bottom" is what you want. Psychologically though, if the alternative is to delay investing entirely due to fear of "buying at the top", I argue this scheme is certainly preferable.
    – wberry
    Dec 5, 2019 at 16:46
  • 1
    Apologies, that was a typo. It should read "increases your risk of not buying at the bottom". Yes, you are correct, that for an irrational investor, such advice can be worthwhile.
    – JBentley
    Dec 5, 2019 at 16:58

The market is within an epsilon, less than the vig you pay a broker, to being totally random and unpredictable. This was shown in an IEEE paper some 30-40 years go.


A properly functioning market will balance supply and demand and sellers by having prices rise in response to excess demand and fall in response to excess supply. Rising prices will encourage sellers to increase production and encourage buyers to find alternatives, while falling prices will encourage sellers to reduce production and encourage buyers to find new uses.

The same should be true for investment markets as for any others; rising prices should be a red flag to discourage further buying. Unfortunately, some market participants may view rising prices not as a signal to slow down, but view them as a bull would a red flag: as a signal to charge ahead. Such participants will push prices higher, which will in turn encourage more bullish buying, until the supply of bullish investment capital runs out. Once that happens, it will become apparent that the bulls' massive overpayment for their assets has done nothing to increase their fundamental value, and the price would ideally fall to what it would have been if the bulls hadn't been insisting on overpaying for them, but may in practice fall below that as a result of panic selling.

Any stock that's worth anything will give the bearer some stream of dividends or other payouts even if it's never sold. The fundamental value of a stock is the present cash value of that stream. People who buy above that price or sell below that price will amplify price oscillations, and will lose money. Those who buy below that price and sell above it will help stabilize prices, and will make money as a reward for doing so.

What's important is not so much to avoid buying when markets are generally high, but rather to buy stocks when their price is below their fundamental asset value and sell when their price is above that value. If something happens that means a stock is likely to pay out more dividends than had been expected, and the change in price is smaller than the expected change in dividends, that may suggest that one should buy the stock despite the price increase, until the enough people buy the stock that the price increase is in line with the asset-value increase.


There are many ways to invest your money in market. No matter what we do and no matter who we are but one thing what we have to understand in which instrument our money will be invested. I want giving underline "when".

In case arguing with many opinions, I start myself with this idea, MoS (Margin of Safety) Margin of Safety


Mean-reversion itself is a purely statistical phenomenon that follows simply from the fact that if you make a random draw that is close to the tails of a distribution (further from the mean), the likelihood the next draw is even further along the tail is less than the likelihood that the next draw is closer to the mean. That's all it is. It "exists" in the same way that the normal distribution "exists".

In an otherwise perfectly specified market with white noise error, all mean-reversion says is that for a given confidence level, the noise is bounded. Of course, the market isn't perfectly specified - you don't know the mean today, you certainly don't know it tomorrow, and since any reversion to the mean that isn't tied to fundamentals has absolutely no connection to physical processes, you can't "time the market" on mean-reversion. Complicating this further, there is no way to prove that the noise has the same distribution today as it did yesterday, so even though it's in principle bounded, in fact you can't rely on those bounds. In other words, knowing that mean-reversion is a thing offers very little information about what the market will do from one moment to the next.

But to answer your questions anyway:

  1. Returns post-2008 were unusually high in part due to survivorship bias, and in part simply due to the fact that recovery by definition requires accelerated growth. If you look at a chart for the S&P500, you'll see that we're only just getting "back on track" with respect to the implied trajectory pre-2008.
  2. Because mean reversion does exist, and its swings can be wild. But also because "most people" don't know the difference between noise and signal in the market, and so can't tell whether a swing in price was "mean-reversion" or a correction. Therefore they treat everything as mean-reversion, which is just another way of saying that they think that trading stocks is gambling.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .