Actionable insights on equities, fixed-income, macros and personal finance Start 14-Days Free Trial
Actionable investing insights Get Free Trial
Charts & Analysis

An Analysis of SIP Returns on the Sensex for 23 years

I have monthly Sensex data since 1978 (the early part of it was reconstructed, since the Sensex only came to being in the mid-80s). So if you invested in the Sensex every month, how would your returns be?

Since you can have different starting dates, I’ve made the simple assumption that what matters is “rolling” returns. That is, today, your metric would be – what would the Sensex return be if I had invested every month for three years (since Oct 2010), for five years (since Oct 2008) and so on.

And in 2007, you would have thought: How would it be if I had invested every month since 2004 (3 years), 2002 (5 years) and so on.

Here’s the rolling return chart since forever:

Rolling Returns for Sensex SIP

(Source: BSE and my own calculations)

Note that all returns are annualized and use the excel “RATE” formula that assumes the same amount invested every month.

But this obviously has flaws. Firstly the data before 1989 (at least for the 10 year) is suspect, because if the Sensex was devised in 1986, how would you know what stocks to have bought in 1989?

Let’s then assume, for convenience sake, that you started an investment journey in 1990. How would your returns be? And let’s look longer term – the 1 year return is extremely volatile and we’ve matured to a point that we understand 1 year is not long term enough.

Sensex Rolling Returns

The three year SIP return, today, is a negative return. In absolute terms, you would have less, after three years, than you had invested since Sep 2010, if you invested a common amount monthly.

Remember, this chart is not adjusted for inflation.

Even the five year return – at +8.7% – does not quite make the inflation return.

The Increasing SIP

When you think really long term like 10 years, your salaries have changed dramatically. So you can’t invest the same amount over time.

In 1990, maybe Rs. 1,000 a month was around 1/3rd of a mid-level salary! And today, it’s the cost of a meal for four.

So let’s assume you start investing Rs. 1,000 a month and you increase the monthly investment by 10% per year. You will now be investing around 9,000 rupees every month. What would your returns look like?

Let’s also add another asset class: Gold. And then, compare it the the “discounted” value of your total investment – meaning a growing line showing a 10% per year return. Instead of percentages, we’ll look at actual figures.

Increasing SIP: Gold and Sensex

The amount invested, at a growth of 10%, becomes Rs. 26.5 lakh (2.65 million), over the last 23 years. Ignore what the absolute amount was – it is irrelevant because purchasing power was different. The 10% a year number is the same as your increase in SIP every year, so it’s a relatively useful estimate.

Investing in the Sensex has dipped below your “cost” (that is, if you assumed 10% return was your bare minimum, a fact in the 90s) at least twice – in the 2001-2003 period, and then briefly in 2008. Even now, the premium of the Sensex over the discounted growth in the cash invested, is not very attractive.

Gold has done phenomenally since 2010.

It is a much more difficult exercise to find out rolling returns in such a scenario. (I have to write a program to do this) But that would tells us how doing an “increasing” SIP would have done, at any point in the last 10 years, and thus, where we stand now relatively.

Impact: Some may choose to view this as the lower point for the sensex which has to catch up to the “mean”. Others may say, okay, the 10 year rolling SIP return is still great – why not keep that as a longer term measure.

And yet others will say, this next decade is the time for gold.

I don’t wish to draw conclusions. Except that purely indexed equity returns may not be as attractive as one thinks, even in the ultra-long term.

  • lohit says:

    >I don’t wish to draw conclusions. Except that purely indexed equity returns may not be as attractive as one thinks, even in the ultra-long term.
    Agree, but the comparison has to be relative. ‘Not attractive’ compared to what?
    One cannot expect more than 4-5% in real returns for a no-effort indexing policy. Want more, then work harder on the investments. Research and then pick stocks/derivatives/metals etc.

    • Compared to Gold, for one. Compared to Actively managed mutual funds, for another. Compared to pure debt, for a third. I will have to work to get real 3 year and 5 year rolling returns on these things!

      • lohit says:

        Which large cap actively managed fund has given these kind of SIP returns over a long period (net of expenses). ?
        As Alok comments below, SIP is just one strategy. There are many more. The reason SIP is widely advocated is that it fits in with the average persons income flow. You get a monthly salary, a part of that gets invested without trying to time the market.

        • A heck of a lot of them 🙂 HDFC Equity, Franklin Bluechip, HDFC Top 200, Magnum Equity and a ton of others. Some of them have done over 20% in the last 10 years.

  • Sidman says:

    Brilliant analysis, this. Thanks for sharing. I’m a huge believer in SIPs, but over the last three years, I’ve been reconsidering that passive approach.

  • lohit says:

    Also in India, long term gains tax on equities is currently zero. Also, MF companies can reinvest gains without having to declare them as distributions and then having them taxed at the customers hands like in the US. That’s an added incentive.

  • SK says:

    Dude, too good! Even though it’s too late for me to start investing, have started thinking recently. Don’t know where to start or which market to invest in.

  • Jaideep says:

    Deepak – over 10 rolling years, the return at 16.0% is substantially higher than the 10% discount rate, especially when we compound. Considering this, why is the difference in the 3rd table (increasing SIP) between the 10% discount rate and Sensex marginal at only 14% over 23 years? Is it due to marginal returns in the last 3 years? Also, has the dividend yield ~ 1.5% been considered in the returns?

  • Alok says:

    I think the period since 2008, or to be more precise, since 2010, is similar to (though not exactly) the period between 1992 to 2003. As per morgan stanley, we are in a similar stage as the late 1990s. If we can analyse the performance of SIPs in the late 1990s, that can give us some clue regarding what to expect in the medium term. There is no holy grail in investing and SIP as an investing strategy is not an exception to this either. Every strategy has its ups and downs and it is during such down periods that conviction is tested the most. I am not an advocate of SIP or any such strategy. Either one can have the ability to know market behaviour in advance and change investing strategies accordingly, if possible. or one should have faith in a strategy and take such down periods as a passing phase. One must understand that the underlying instrument for SIP strategy is ultimately equities and if the underlying doesn’t do well, strategies by themselves cannot give high returns. What strategies can however do is – reduce risk during such times, which i think SIP strategy would have done over last 3 or 5 years.

  • Arun says:

    Probably naive question here: What about dividends? Are they factored in as as returns or re-investments?

  • dheeraj says:

    Converting these return numbers into some sort of a probability distribution gives us more meaningful information.
    For instance, an estimate of the mean and the range (or variance) of rolling returns (1 year, 3 year. 5 year or even more) over the past 20-25 years would give us meaningful information aiding us in decision making.
    One could even test hypotheses at different confidence intervals and thus take informed decisions. (yes i know these are statistical techniques that not everybody may be conversant with – nevertheless they are useful)

  • DJ says:
    DJI is like the SENSEX. But, I’m sure middle cap or small caps have done better both in the US and in India. And, not necessarily relying on active management, but just going beyond the index helps. SENSEX and DJI are also actively managed in some sense (using a rule) and the rule is limiting since only big companies get added to the index, which by definition means that the high growth period is beyond them. In that context, DJI outperforming gold is a pretty attractive proposition.

  • VK says:

    Nice analysis and report. Appears as if 2002-2004 was a really bad year. No matter your investment duration (for that period), the best you did was make no profit! Why is that?