Is there a “best time” to invest in the FTSE?
The latest Two Guys & Your Money podcast recently arrived in my itunes directory! Awesome – I always look forward to listening to these guys. However, the topic of this podcast really interested me this week. It boiled down to the fact that you should never follow the herd when investing. This was in line with the age old advice from Warren Buffett saying: “Be Fearful When Others Are Greedy and Greedy When Others Are Fearful”.
However, I have also heard the advice that “you should never try to catch a falling knife” and “you should never go against the flow of the market”.
For me, these two sets of advice are somewhat contradictory.
I therefore want to address the question: “when is the best time to invest in the market”? And I did not know which of these schools of thought was right. So, as I usually do, I did some data based research, and have concluded that (although there are a couple of times to avoid) there isn’t really a “best” time to invest in the stock market.
FTSE 250 performance
This analysis uses the raw data for the monthly closing position for every month between December 1985 and September 2013. When analysing the figures below, it is good to keep in mind that the monthly average performance of the FTSE 250 over this period is 0.7% per month (not including dividends).
Month n compared to month n-1
For example, this looks at the performance of the market in one month (y axis) based on performance in the preceding month. For example, if January decreased by 5%, what does February do on average? And so on:
- If the market fell in the previous month by more than 8%, the performance in the next month is, on average, negative.
- If the market rose in the previous month by more than 14%, the performance in the next month is, on average, negative.
- Anything in between -8% and +14% leads to a positive average return in the following month.
- The best returns in the following month are noted when the previous month performance is between- 7% and -3% or +7% and +13%.
The quick conclusion we can draw here is that markets generally increase in month n, independent of the performance in month n-1. However, there is an exception whereby markets generally perform badly following an extremely volatile month.
This leads me to ask whether, following a month of extreme volatility (and where we are already invested in the market), we should consider selling and buying back into the market the month after?
I would always say never make decisions in the market based on short term factors. However, the data and the maths perhaps prove this wrong.
Imagine we have a £20,000 position and it loses 9% in a month. The remaining value is £18,200 and this loses, on average, 2.1% in the following month.
Selling out and re-buying (based on my current broker’s fees) would cost £5.75 x 2 in trading fees and £54.60 (0.3%) in bid-offer spread, totaling £66.10. However, on average, we would lose £382.20 (2.1%), on average, by holding our position in the market. Therefore, it is significantly cheaper to sell-out and re-buy in one month in this example.
The problem with following this technique is the lack of data. There are only 18 instances in the past 27 years where the market has fallen by over 8% (the value required to make this worthwhile). Therefore, the data is too limited to draw exact conclusion on this matter.
That said, we have a fairly large margin in our example above (1.8%), so it could still be a good idea to follow this sell and rebuy plan.
Month n plus month n+1, compared to month n-1
Let’s increase the time period and base our results on the prior one month’s performance, but now look at the total performance of the following two months:
- No average negative performance over 2 months, independent of the prior month performance.
- No real trend noted.
Does this provide us with any new information regarding the conclusion reached above? Not really. It tells us that the performance 2 months later, on average, will be back to a positive number independent of the prior month performance. Does this suggest we shouldn’t sell and re-buy? Nope.
In our prior example, our previous month data of 9% loss provides a positive return over the following two months of 0.1%. Therefore, this seems clearly better than incurring the costs and staying out of the market for two months.
However, this is not our actual comparison. Above, we said we sold out of the market, which cost us £66.10. We then bought back ONE month later, after the market had on average decreased by 2.1%. If we then look at the ONE month performance for the following month, when that month decreases by 2.1%, we see a positive gain of 1.4% in the next month.
So, our total position after the sell and re-buy is £66.10 loss, followed by a 1.4% gain, gives us £18,387.77 after the two months.
If we had held our position after the initial 9% decrease, we would have seen a 0.1% increase on £18,200 giving us £18,218.20 position.
The difference is 0.9%. Worth having for one month!!
Month n, compared to month n-1 plus month n-2
What happens if we look further into the past?
Similar result to just looking at one month. The values at the very extremities lead to a worse return that in the middle. Investors do not like volatile markets.
Month n plus month n+1, compared to month n-1 plus month n-2
And the same again compared to the 2 months following one month. No new conclusions here.
Finally, let’s look at quarterly performance based on the prior quarter:
Quarter n, compared to quarter n-1
It seems this time, over the slightly longer term, avoiding extremities is not necessarily a bad thing.
I would intuitively think that after 3 months of losses totaling over 14%, we would be at a good buying point. However, although performance is positive for this area, it is only around the monthly average.
It seems that the place to buy, and I would not have intuitively thought this, is actually when the market is on a run. This seems to suggest “buy when others are greedy”, which goes against one of Mr. Buffett’s most famous sayings.
However, due to the limited data for 3 month runs over 16% (there have only been 14 instances in 27 years), I’m not going to place any concrete conclusions on this.
So, when is the best time to invest in the FTSE?
For me, the best time to invest in the market is when it’s pretty bloody boring (average recent monthly gains or losses of about 2-3% either way). However, gains over one month, two months and three months are better indicators of positive performance in the following month than the equivalent losses.
There does seem to be some optimal times to invest (for example, buying into the market when it is on a 3 month run bull run), but this has been derived from a very small data set. As such, I would not say that there is one specifically good time to buy.
However, I would argue that there are times to avoid investing, specifically when the market fell in the previous month by over 8% or increased by over 14%. Investors like stability and whilst these types of movements do provide opportunity through greater volatility, the following months generally provide below average, and in fact negative, returns. So actually, we should not always follow other investors, and we should not always do the opposite.
The overall conclusion is that we should always buy the market, regardless of what “market opinion” is, EXCEPT when the market is extremely volatile over the short-term.
However, the interesting thing that this research has told me, which I haven’t considered before, is that it may be statistically optimal (although dependent on the amount invested and the relative fees involved) to sell out of the market for one month only and re-buy when the market demonstrates extreme levels of volatility (either positive or negative). This assumption is based on a relatively small volume of data (335 months), but the margins are so great that this makes up for the limited data set.
There are some other things to consider when determining the optimal time to start investing.
These are, with the biggest mistakes first:
1) Not being in the market
2) Being under-diversified by buying a narrow range of individual stocks
3) Being under-diversified by buying on a limited number of dates
Number 1 is self explanatory. Monthly average returns of 0.7% not including dividends is a very good reason to be in the market at any stage.
Number 2 poses the risk of over-exposure to individual companies failing. Imagine we bought HMV, Blockbuster, Connaught, etc etc over the past few years. We certainly wouldn’t be averaging that 0.7% now. We can avoid this by buying ETFs to track the market performance as a whole.
Number 3 is less clear cut. Firstly, we want to diversify our entry points. Imagine you bought everything on 1st May 2007. The good news is that the market would be still be up 4% per year between May 2007 and October 2013.
However, if you had bought at the start of December 2008, you would have had 23% annual returns. Therefore, to reduce our variance, spreading the purchasing time across several months and years is advisable.
However, you have to balance this out with the increase in fees that you would incur by investing more frequently.
Say you have £12,000 to invest in a year (ISA limit for stocks & shares ISAs is £11,560 in the current year). In my opinion, it would be much better to buy £4,000 in a FTSE 250 ETF every 4 months than £500 twice a month. Although this exposes you to greater timing risk, we can see in the “2 month” and “3 month” averages above that this impact is not huge, especially when compared to the fees incurred:
Conclusion – the best time to invest
So, what mantra or quote should I be following to determine the best time to invest in the markets?
Well, my friend, none. Don’t follow the herd. Don’t be greedy. Don’t try to guess the flow of the market. Don’t try to catch falling knives or sell at record highs, etc. Just buy at all times other than in very volatile short-term markets.