You may have heard statements like, ‘equity investing is for the long-term and not for the short-term’. Exactly how long is ‘long’ and how short is ‘short’?Here is how short-term and long-term are defined.
This post assumes knowledge of CAGR and standard deviation. If you are not familiar with these, but interested to know more, do consider watching these short videos:
In both these videos , I had used the annual returns of ICICI Pru Top 100 fund.
The green line represents the arithmetic average (what we calculate usually), 29.3%. The blue lines represent deviation of each annual return from the average. The average of these deviations defined so that it is always positive is the standard deviation: 38.6%.
So this data is usually reported as average = 29.3% +/- 38.6%. This means (to those who familiar with the normal distribution) that each annual return has 68% chance of being in that return window. This is both simple as well as simplistic. However, it is a good way to understand investment risk.
the standard deviation is a measure of uncertainty in the average return. If the uncertainty is larger than the average return itself, then the average has little value. This is the key to separating long-term from the short-term.
To know more about the normal distribution, you could consult What Return Can I Expect From Equity Over the Long term? Part 2
Suppose I ask myself, “what return (CAGR) can I expect if I invest in equity (in this fund in particular) over the next 5 years?”. To answer this, let us look at the CAGR of every 5-year window in the past with the above annual returns data.
CAGR from 1st Jan 2003 to 31st Dec 2007: 47%
CAGR from 1st Jan 2004 to 31st Dec 2008: 13%
CAGR from 1st Jan 2005 to 31st Dec 2009: 23% and so on as shown above.
Notice that the 5-year window ‘rolls over’.
The arithmetic average of all these 5-years CAGRs is 17.5% with a standard deviation of 13.7%.
If I report 17.5% as the average CAGR in the past, that looks terrific. However, if I report that the uncertainty associated is 13.7% that return does not look so hot. Standard deviation is a measure of volatility and potential investment risk.
Now let us look at a truly incredulous graph. Suppose I had annual returns of the Sensex total returns index ((dividends included) from 1979 and 2013 and calculated the CAGR of every 3,5,7,10,12,15,20 and 25 year periods possible and calculated the average CAGR, it would like this
If I gave out only this information then it looks as if the average CAGR over 3-year investment periods is as high as that over 25-years!
The key information missing is the standard deviation. That is we must also report how much individual 3-year CAGR can deviation from the average.
This to me is the defining graph for investment risk. Notice how the standard deviation is extremely high, double-digit for up to 7-years and then drops down. This means for up to 7 years, there can be significant volatility in equity returns.
My long-term starts at 10 years and I would prefer 15 years if possible.
My short-term is 0-5 years and I will not touch equity for such durations.
My intermediate-term is 5-10 years with about 10-30% equity exposure.
My long-term is 10+ years and anywhere between 50-70% equity exposure with the recognition that my entire portfolio could fall in value by 50-70% if there is a market crash. Read more: Asset allocation for long-term goals
Here is a step-by-step guide to decide asset allocation for a financial goal
These are of course my definitions of short-term and long-terms. Definitions are personal, but I think ought to be backed by some kind of reasoning. Using the standard deviation is one way to do it. Would be delighted to know more about your definitions.
Caution: Do not assume that long-term equity investing will always be successful! The nature of the above graphs will change over time.