Beware - Numbers can be misleading!

 "XYZ scheme is a consistent performer. Our fund has been a consistent performer. We only recommend consistent performers."

We keep hearing (or reading about) many such statements. The natural reaction is to put our money in such schemes. But, have we ever wondered what consistency is? Well, we often take certain things for granted. When someone mentions "consistency," one would mean it, isn't it?

We chanced upon one article in a leading newspaper about a fund. The author declared that the fund was a consistent performer, and obviously, recommended investment in it. Knowing the reputation of the newspaper in question, I am sure the fund's corpus would have swelled.

One thing caught my attention - the basis of declaring the fund as a consistent performer, in fact, consistent outperformer. The fund had beaten its benchmark for various time period, short term, medium term, and long term. Let us reproduce the performance numbers for better understanding:

Period	Fund	Benchmark
1 year	25%	9%
3 years	17%	13%
5 years	9%	8%

 

 

 

 

(Although we have not mentioned the name of the fund, we would still like to remind our readers that "past performance may not be sustained in future.")

As can be seen from the table, the fund has outperformed the benchmark over various time horizons, viz., 1 year, 3 years, and 5 years. This is enough for many to consider the fund as a very good investment option.

Warren Buffett has once said, "A public opinion poll is no substitute for thought." We decided to follow the Oracle in the above example and dug deeper. A careful look suggested that someone hurried to the conclusion. The time periods need to be non-overlapping in order to decide if something was consistent, whereas in the example above, there were overlaps. Last 1 year is part of last 3 years and that is part of last 5 years. Can we find out if there is any impact of the overlap on the conclusion?

Let us reconstruct the NAV based on the performance numbers given in the table above. Assume that the NAV of both, the fund and the benchmark, is Rs. 100 at present. Considering that the fund returned 25% in last one year, its NAV would be Rs. 80 a year ago. 25% return over a year will take an investment of Rs. 80 to Rs. 100. Similarly, investment of Rs. 91.74 would become Rs. 100 over a year if the return on investment is 9%.

We arrived at the values for both the fund and the benchmark in this fashion.

The values are mentioned in the table below:

NAV	Fund	Benchmark
Today	Rs 100.00	Rs 100.00
1 year ago	Rs 80.00	Rs 91.74
3 years ago	Rs 91.51	Rs 103.06
5 years ago	Rs 64.99	Rs 68.06

 

 

 

 

 

 

Based on the above data, let us break the last 5 years into 3 distinct and non-overlapping periods, viz., last 1 year, middle 2 years, and the first 2 years. If the period under consideration is last 5 years as on 31-Dec-2011, the three non-overlapping periods mentioned here would be as under:

Last 1 year = 1 year completing on 31-Dec-2011

Middle 2 years = 2 years completing on 31-Dec-2010

First 2 years = 2 years completing on 31-Dec-2008

We can find what returns the fund and the benchmark gave for these periods. And the results are stunning:

Returns
Period	Fund	Benchmark
1 year	25%	9%
Middle 2 years	-7%	-6%
First 2 years	19%	23%

 

 

 

 

 

Now this table tells a completely different story as compared to the first one.
To look at this from another angle, let us say we started with investment of Rs. 100 and see the value of investment at the end of several periods.

Value of investment after the period if invested Rs. 100 at the beginning of period:

Period	Fund	Benchmark
1 year	Rs. 125.00	Rs. 109.00
Middle 2 years	Rs. 86.49	Rs. 88.36
First 2 years	Rs. 118.81	Rs. 151.29

 

 

 

 

 

The fund underperformed in two out of the three periods discussed here. It was only in the last one year that the fund erased the entire history and the fund looks like a long-term hero. One can only conclude from this that (assuming that benchmarking is correct) the fund outperformed the benchmark by miles in last one year. It does not mean whether the fund is good or bad. The other thing we can definitely and conclusively suggest is that there is no consistent outperformance.

We suffer from what is known as "availability bias." We have a habit of taking the available data and jump to conclusion. In evaluating performance, we seldom break the performance down and simply try to interpret the available data. It is important to look beyond what is obvious.

This article only talks about looking at consistency and hence we have not touched upon a few other items:

1. Whether the benchmark is relevant
2. The fact that past performance may not be sustained in future must be remembered before investing in anything by simply looking at the past.

One more thing before we stop. It is important to mention here that mutual fund is arguably the most transparent investment avenue, and hence too much can be written about mutual funds and gets written, too. However, the principle of consistency that we have mentioned here is applicable to each and every investment option that one may consider.

Happy investing.

- Amit Trivedi

The author runs Karmayog Knowledge Academy. The views expressed are his personal views. He can be reached at amit@karmayog-knowledge.com.