May 20, 2012

The problem with historical returns

What is the most commonly heard refrain about the stock market these days?

My guess is that a lot of people now believe that the stock market is a nasty, volatile place where a serious investor cannot make any money. It is a place for gamblers, traders and at best for the short term investor. It is not the place where you invest your retirement money.

One cannot blame the common man for this view. The recent history of the stock market has only re-enforced the above viewpoint. The problem however is that recent history is a poor guide to the stock market or as a matter of fact for any asset returns.

Some historical numbers
Let’s look at some numbers.

The sensex went up roughly from 1300 levels in 1991 to 4000 in 2000. This gives us an average annual return in the 10-12% range. The sensex then rose from around 4000 to 20000 in the next ten years, returning around 17.5% per annum.

These returns are very impressive and also completely meaningless. These numbers hide more than they reveal. These numbers hide the fact the stock market returns are lumpy and do not come in smooth even intervals. None one made an even 17.5% per annum return during the period from Dec 2000 to Dec 2010.

Let’s break down this period as follows

Dec 2000 – Dec 2003: Index went from 3973 to 5838 (13 % per annum)
Dec 2003- Dec 2007: 5838 to 20286 (36% per annum!!)
Dec 2007 – Dec 2010: 20286 to 20509 (around 0.4% per annum)

As you can see, the returns have been lumpy and were concentrated in the 2003-2007 period.

How does the common investor behave?
Imagine an investor in 2007, who has always invested in fixed deposits, gold or real estate. He has been watching the stock market for the last 4 years and has seen the stock market rise by 300%. He is watching his friends and relatives get rich. At the same time, every time he or she visits the bank, the nice personal banker tries to push the hot mutual funds of the day by showing the fantastic returns of these funds for the last 3 years.

If you were looking at the data in 2007, it looked fantastic no matter how you sliced and diced it. The 1, 3, 5 and 10 or 20 year returns looked good.

So let’s say you got taken by the historical returns and went and bought a whole bunch of mutual funds and stocks. What happened after that?
Dec 2007 – Dec 2011: 20286 to 15454 (- 5% per annum for next 4 years)

Ouch!!!

What is the general perception now?
I have been reading quite a bit of the analysis that the stock market is a bad place to invest. Even if you are a long term investor and were invested for the last 3, 5 or 10 years, other asset classes such as fixed deposits would have beaten the stock market at a much lower risk.

I find this argument shallow and intellectually lazy.

The problem with this argument is that the person making this argument is doing data mining. He is slicing the data in such a way that it just proves his point and does not really highlight the main point about the markets

So what are the main points?
I would say there are several points worth remembering
  1. The stock market is a volatile place and returns come un-evenly. As you saw from the data above,  past returns have not been smooth fixed deposit type returns, but lumped in short periods of time.
  2. Valuations matter! If you buy at high valuations (dec 2007) and sell at the time of low valuations (say Dec 2009), you will lose money. Period!
  3. The stock market is a risky place. There will be long periods of time where you will not make money or even loose money. At a point when everyone is pessimistic or has given up, the stock market has a tendency to turn and surprise everyone. The same holds true at market peaks too.
Other asset classes
Let’s look briefly at some other asset classes.

Gold (all prices in dollars per troy ounce)
1971- 1981: 40 – 460 (25% per annum)
1981 – 1991: 460 – 362 (-2 % per annum)
1991 – 2001: 362 – 271 (-3 % per annum)
2001 – 2011: 271 – 1571 (19% per annum)

As you can see from the above numbers, gold seems to have followed a similar trajectory. There have been periods of high returns, followed by long periods of dismal returns (40 year returns have been around 9.5% per annum)

I don’t even consider gold as an investment as it does not generate any cash flow and is merely an insurance against armageddon or end of the world scenario. But I think I am in the absolute minority, considering the fact that Indians are the largest buyers of gold and absolutely love this metal. So in the end, one cannot really put a price on love!!

I don’t have the numbers for real estate, but anecdotally real estate has displayed a similar pattern. The returns were poor from 1993 to around 2003. The major gains came from 2003 to around 2008 and now the real estate market has slowed down considerably.

You will definitely find examples, where someone purchased a piece of land outside the city and was able to get 10X his or her investment. However a single multi-bagger is not representative of an entire asset class.  That’s like saying that as Hawkins cooker went up by around 1600% in the last 5 year, the entire stock market should also have done well.

The curse of past returns
I am not optimistic that the general, un-informed investor is going to change any time soon. The majority of investors are hard working, middle class people with busy lives. Investing and  the stock market is the last thing on their mind. The time when the market does catch their attention, is when it has gone up considerably. As a result, most of the retail investors end up entering the market at precisely the wrong time.

Past returns are a good starting point to evaluate the long terms returns of an asset class. However these returns are not written in stone. The best approach to evaluate the likely (not guaranteed) returns one will make, is to calculate the expected returns at any point of time and make buy or sell decisions accordingly. The topic of expected returns is however a much more complex topic, and possibly one for a future post.

21 comments:

Anonymous said...

Hi Rohit,
Great post and very timely.

AMCs/ Brokers/ Expert always talk about great long term equity returns without based on a suitable time period. Lot's of people right now are saying 5 years of nothingness in equity. Death of equity investing etc (without talking about the high valuations in 2007)

Have a request. Could you sometime look at say a 5 or 10 year returns from equity based on valuation at the time of investing (For different P/E and P/B bands)

For example: If you invested at P/E of 18, avergage 5 year returns were, whereas if you invested at P/E of 20, the returns were YY%

Thanks
Kamlesh

Dhwanil said...

Hi Rohit,

Excellent post. Inspite of understanding all the points here, I am unable to convince some of my closest relatives that market is the best place to invest at right time for long horizons! As you very rightly pointed out, it is lumpiness of returns that is not understood and most retail investors exit the market after 2-3years of mediocre/negative return, precisely wrong time to exit/stop investing. I hope more and more retail investors read this.

Best Regards
Dhwanil Desai
http://www.valueinvestinginpractice.blogspot.in/

Nitin Jain said...

As always great analysis ! Even the publicly available data can be tweaked so as to paint a desired picture. We the retain investors end up investing in few stocks, few MFs or 1 or 2 real estates in the life time, so instead of following the advise of 'see I told you' friends and advisers, who also might be right in their own might, one has to focus on his/her own universe !

Your analysis proves the fact that Simple Linear projections simply are useless but interestingly all the Brokers and Advisers on TV and Free media keep giving advise based on these simple linear projections.

mypursuitofperfection said...

Good one Rohit.

In other words, this post tells us "Keep the Faith". :-) Of course keeping all the points in mind.

thanks,

Vikas

Anil Kumar Tulsiram said...

Excellent

mkd said...

Rohit,

Excellent article as always.

I am with you on gold. If its a minority, then it's at least a minority of two.

Niraj said...

Rohit I believe all assets class is good or bad and no asset class is better than others. Stock market history in its current form is not more than 130 years old and even if you consider US market there is quite a long period in which stock market has not offerred good return for example: between 1929- 1949 and 1965- 1981 and all these are fairly long period of times. Globally from 1981-1907 is one of the best return offered by stock market including indian market and partly reasons for the same may be due to increasing reach of technology which make investment in share market easier than investment in fixed deposit in nearby banks. Also at any point of time anywhere only 20 % of the companies share are of investment quality and likely to do well in future whose prediction is not easy even for a experienced investor. You are a value investor,about a year back i have written a article about Warren Buffet investment return in my blog http://sensexvalue.blogspot.com which shows even his return between 2000-10 is not impressive.

Unknown said...

Rohit,
I differ from you & other intellectuals on Gold Prices.
Can you make the same chart in simple indian rupees? Just like you buy or sell stock in rupees, you will buy / sell gold in rupees only.

Whenever gold prices decreases (in dollar terms), Dollar appreciates vs rupee. So the cost of gold in Rupees will be increasing.

Anonymous said...

Hi Rohit,

great article as usual-thanks.

Are we missing to add divident yield while counting returns?

At present sensex divident yield is 1.82% and if we substract 0.8% expanse ratio of ETF effective yield would be 1%.

Person who has bought index at 2000, 20 years back would have current yield of 8%.

If we assume dividend yield in 1992 was 0(because of high PE than of Harshad Mehta era) even then average yield would be 4.5% per year over 20 years.

If we add this 4.5% to capital gains equity return would be much higher and could beat almost all asset classes over that time.

Regards,

Ashok

ValueFactors said...

Hi

I agree valuations matter but the probelm as I see it, is how to decided in the moment whether valuations are high or low. There always seems conflciting evidence that supports they are too high or too low.

Interesting blog by the way!

Regards
Value Factors

Vignesh said...

Hi Rohit,
your Views on Infosys. Wonderfull compnay having some temporary issues... Also feel that the valuation has come to a reasonable level... Pls let me know your thoughts on infosys....

Rohit Chauhan said...

Hi kamlesh
i have done some rough analysis in the past and even posted on the blog. any time the trailing PE dropped below 11-12, the future returns have been good. however i have never tried to be anymore precise than that
rgds
rohit

Rohit Chauhan said...

Hi dhwanil
one can never convince people to invest in the bear market. if it was easy to convice then we would not have current situation...personally i dont even try :)

rgds
rohit

Rohit Chauhan said...

Hi nitin
general media cannot look beyond their nose and most advisors need to provide instanteous results ...i personally feel it is very difficult to make a living as a short term advisor

rgds
rohit

Rohit Chauhan said...

Hi vikas
true ...though lets see how much the faith will be tested. i have a feel , it may be tested a lot

rgds
rohit

Rohit Chauhan said...

Hi mkd
thats makes it two of us ...well my wife and mother are not in the camp :)

i buy gold for them as an expense and never consider it as an expense :) i have suggested to them that we can sell the gold and buy stocks and the look i got :) will never try it again

rgds
rohit

Rohit Chauhan said...

Hi niraj
true ..point taken. not asset class is good or bad. it is the price which is important.

however is gold an asset class ? does it generate any cash flow ? if you buy 100 gm gold ..it will still be 100 gm after 10 yrs.

if you buy real estate, atleast there is rental income.

i would define an asset as something which generates cash flows. gold, art etc are all speculative assets in my opionion

ofcourse almost no one believes other than me :)

rgds
rohit

Rohit Chauhan said...

Hi unknown
you are mixing two different things. rupee dollar rate and gold prices

rupee appreciated from Mid 2009 to early 2011 even as gold was going up.

now due to the europe issues, we have a flight to dollar and gold for safety issues and hence the concurrent rise

there have long period when both these variables have moved in different directions

rgds
rohit

Rohit Chauhan said...

hi ashok
you are right ...the index returns are excluding dividends. if one add its back then the returns are even higher

and that is why is it a productive asset ..it yields ever increasing cash flows
rgds
rohit

Rohit Chauhan said...

Hi valuefactors

well deciding valuation is really an art form. i would say there are levels at which one can be more confidence that it either over value (say above 25+ PE) or undervalued (say below 10).

again valuation is always in a context ..one has to evaluate multiple factors to make a decision. that is why i think it is much easier to value a company than try to come up with a decision for the entire market as a whole

rgds
rohit

Rohit Chauhan said...

Hi vignesh
i am watching the company too, but i think the company is still stuck in an old mode of working. they are chasing margins which are difficult to get

all this story about premium service and infosys 3.0 ...is just story ..to put it politely

i have worked in this industry for too long and have yet to see anyone sell premium service in volumes. again indian IT companies have no IP at all unlike an IBM

i would not look at past profits and valuations to make a decision.

rgds
rohit