Should you invest at market highs?

Indian markets have been making new highs lately. Naturally, the most frequent question doing the rounds is:

Should you invest at market highs?

The question tends to be phrased as "Isn't it just plain dumb to be investing at market highs? Should I a) stop investing more and b) sell everything and buy after the inevitable crash?"

The idea of selling near the top and buying everything back much cheaper is attractive. There's not much evidence to say it works.

The chart shows the Nifty since 1999. The shaded regions represent the times it made new highs.

Between July 1999 and Nov 2022, the Nifty closed at a high on 713 days out of 5,833 days, roughly 12% of all trading days.

The question we're trying to answer is: Whether investing at highs is worse than investing on any randomly picked day.

We know over the long term, markets trend up, so given time, buying the index at any level would have delivered positive returns. Since the Nifty is currently close to a fresh high, any investment made in the past, whether or not at past highs, is positive as of now. Note this applies to the specific Nifty index only and not individual stocks.

But the spirit of the question we're answering is about being right or wrong on a shorter time horizon. The magic horizon most seem to (wrongly) consider "long enough to decide whether something is working" seems to be one year. So we'll work with that.

The chart shows the distribution of 1 year returns if you invested every day from July 1999 to Nov 2022.

Quick things to note

Nifty 1-year returns are a wide range: from -60% (invested in Jan 2008) to +105% (Apr 2003) over any 1-year time-frame
27% of the time, a little under a third of the time, Nifty goes negative over one year. Conversely, 73% of the time, the Nifty is higher a year later.
The "typical" Nifty return of 12% only happens 11% of the time (the tallest bar in the chart above). This means 9 out of 10 times, the Nifty return falls outside that often-quoted 10-15% range. This means plugging in that 12% assumption for short horizons like 3 and 5 years will likely be way off.

What happens to money invested on days when the Nifty hits all-time highs?

The chart shows return distribution when investing any day that the Nifty is up to 0.5% below an all-time high.

First, our sample shrinks to just over 700 days, i.e. the number of days Nifty closed close to a high.

Investing at all-time highs is not as outlandish as it seems.

Overall, 28% of the time, 1-year returns from investing at an all-time high have been negative, almost the same as if you picked any random day to invest. This is intuitively lower than what most would expect.
The shape of the distribution has differences: Investing on any given day gives you a 0.8% chance of seeing a catastrophic loss of more than half your invested amount (that bright red tail on the left). That probability increases to 2% if you invest at all-time highs.
The other difference is we don't see extreme positive returns of 70% and above over the year after investing at highs. This makes sense since the massive years tend to follow significant corrections.
Interestingly, the long-term Nifty average return of between 10 and 15% is more frequent when investing at highs, nearly 17% of the time versus 11% otherwise. The most frequent return from investing at highs is a solid 10 to 15% return.

Investing every day means investing at highs too. So what if we invested every day except when the Nifty was making highs?

By choosing not to invest at highs, we reduce the long tail of extreme negative returns by a little, from 0.8% to 0.6% of the time and a 4.3% probability of extreme positive returns. The overall improvement to expected returns from choosing not to invest at highs is modest at best.

A simplified summary of the above three charts together

Are index lifetime highs a particularly bad time to deploy capital into the index? History says no. The chances of negative returns over 1 year are only marginally higher than investing at any other time while the base case remains positive returns.

Since most of us deploy small chunks of capital over the course of a couple of decades, the significance of the timing of any one such set is even less critical.

And if highs are not notably worse times to add capital, selling existing investments makes even less sense. The first chart shows the shaded regions where highs were made, and selling at the first sign of a high would mean experiencing gut-wrenching regret as the index continues making new highs for a while.

How about over three years? We plot the same returns distribution but consider holding period of 3 years.

Notice how increasing investment timeframe from 1 to 3 years reduces the "width" of the distribution, which means the range of probable outcomes from investing for 3 years is narrower than any one year. The same pattern applies to the range of outcomes from active mutual funds, as the time frames increase, the range of returns from the best to the worst narrow.
The peak of the distribution of 10 - 15% annualised return now occurs nearly a quarter of the time, and more than half the time, the returns are in the 0 - 15% range
The probability of negative returns drops by 2/3rd by investing for 3 years: from 27% to 9%. When investing at highs however, that probability of negativecomes down from 28% to 18%, still a reduction just not that big

One way to look at the 1 and 3-year outcomes is that all-time highs don't matter for SIP investments so for someone looking to deploy a large lumpsum, just convert them into SIPs and deploy gradually to avoid the load of having to decide when is a good time.

Highs are not a better or worse time to invest, but how about lows?

Waiting for lifetime lows is an unviable strategy, given the index's upward drift. You'd just be sitting on the sidelines for almost the entire time. So as an exercise, we considered what happens to capital deployed closer to 52-week index lows.

The chart shows the investment outcome when the index was within 5% of its 52-week low.

The entire distribution shifts to the right with no extreme negative returns and more extreme positive returns. Surprisingly, the probability of negative returns still is a significant 21%, but down from 27% at any other time. The most frequent annual return continues to be 10-15%. The median return, which was 11.4% for investing any day, moves up to 13.4% when investing near 52-week lows.

But note this would have been implementable on only 390 days over 22 years, less than 10% of the time. You have to stay out of any investing for 90% of the time, making an "enter only at 52-week low" strategy unrealistic.

Bottomline: Investing at all-time highs is not dramatically different from investing at any other time. The chances of negative returns one year hence move up ever so slightly, but the most frequent outcome is still positive. On a 3-year timeframe there are bigger differences in potential outcomes but still a reduction in probability of negative outcomes from investing at highs and not at highs.

The investor building a corpus for the long-term is better off continuing to invest irrespective of whether the index is making highs or lows and increasing the amount to deploy when she can, and not based on what the index is doing.

Let us know what you think. We're on twitter @CalmInvestor and @Capitalmind_in. If you'd like to talk to us about our PMS, get started here

Should you invest at market highs?Should you invest at market highs?

Related Posts