We always think that it is best to invest when markets are down. But how often are we able to do it? Not regularly.
Why?
There are 2 reasons.
First – The markets don’t go down regularly.
Second – When they do go down, we don’t have the guts to go out there and invest. And that is because, either we don’t have the money to invest AND / OR we don’t know where to invest.
And frankly speaking, both are genuine problems. It happens to me, you or almost anyone I know.
I was just thinking about this whole scenario of investors feeling handicapped when markets go down. They want to invest. But don’t know where to invest. And if they do know where to invest, then they don’t have funds to do it.
So is there a solution to this problem?
I think there is. Atleast I thought there was. 🙂
I am not sure though.
I am not sure though.
And if there is a solution, then it is ought to be based on the following two thoughts:
First is that it does not make sense to stay out of markets at any time. I covered this topic a while back at – Why You Should Invest, Even if You can’t Beat the Market.
And Second is an extension of the above thought – it seems wiser to invest ‘more’ when markets are down.
So how do we use these these two (thoughts) in combination to gain the maximum benefit?
First is that it does not make sense to stay out of markets at any time. I covered this topic a while back at – Why You Should Invest, Even if You can’t Beat the Market.
And Second is an extension of the above thought – it seems wiser to invest ‘more’ when markets are down.
So how do we use these these two (thoughts) in combination to gain the maximum benefit?
I came up with a very primitive hypothesis:
Investing regularly (Rs X) & periodically making lumpsum investments* when markets are down, will fetch higher returns than regularly investing double the regular amount (Rs 2X) over a period of 10-15 years in a decent, well diversified mutual fund.
* This lumpsum amount would be an accumulation of Rs X every month, which will be used at one-go when markets are down (at a pre-decided trigger point).
To test this hypothesis, I chose HDFC Top 200 as the fund. The choice of fund was because it is a stable, well known and proven fund. And my guess is that If you had to choose a fund blindfolded, chances are you would choose HDFC Top 200. Ask anyone and almost always, this fund’s name pops up.
So coming back to the analysis, I went ahead and used last 20 years data to prove my hypothesis.
There are 2 scenarios being compared here:
First is where Rs 5000 is invested every month in HDFC Top 200. An equal of Rs 5000 is also parked every month in a Recurring Deposit which I call Market Crash Fund. (I am ignoring the interests accrued for simplicity). The money keeps accumulating in this market crash fund till the time a trigger* point is reached.
Second scenario is where Rs 10,000 is invested in HDFC Top 200 every month. Plain and simple.
*The trigger point for utilizing the market crash fund is when the average monthly NAV of the fund is at more than 20% discount to highest average NAV, which has been achieved in last 3 years. For example, if the fund reached a maximum NAV of Rs 100 in last 3 years, and the current NAV touches 79 (i.e. 21% discount) – then entire money accumulated in Market Crash Fund is used to buy units of Mutual Fund at lower NAVs.
So with these assumptions, lets see what happened.
Starting October 1996, the total money outgo in both scenarios is Rs 21.9 Lacs. In first scenario, Rs 10.95 lac is invested as SIP of Rs 5000 every month whereas rest is invested in lumpsum at regular intervals as and when a trigger point is reached. In second scenario, entire Rs 21.9 lac is invested as SIP of Rs 10,000 every month.
But here comes the surprising part. The returns of above strategies are not in sync with my expectations. My expectation was that the MF+RD combination which shifts money from RD into MF when markets go down, will beat a simple MF SIP of amount equal to MF+RD monthly investment.
But results are completely opposite.
But results are completely opposite.
The MF+RD combination delivers Rs 2.14 crores whereas a simple MF delivers Rs 2.43 Crores.
 |
| Value of Investments in HDFC Top 200 in both scenarios |
In the graph below, the dark blue line indicates the value of Market Crash Fund which accumulates Rs 5000 every month till the trigger point (Current NAV < 80% of Max NAV in last 3 years) is reached. The light blue area graph shows how low is current NAV (in %) with respect to maximum NAV in last 3 years.
 |
| Lump Sum Investment Trigger Points (On Basis of current & last 3 years NAV) |
Big figures indeed for small Rs 22 Lac investment. Isn’t it?
But what about the hypothesis?
How is it that a simple MF SIP of Rs 10,000 a month beats a combination of Rs 5000 a month MF and lump sum investments when markets are down?
Is it something about the assumptions, that went wrong?
Is it that trigger point should have been either more than 20% or less than that?
Is it wrong to split the investment amount into 50%-50% for SIP & RD in first scenario?
Is it because of the extraordinary bull runs of 2003-2008 that this analysis went for a toss?
Is it that HDFC Top 200 has been a stellar fund and any other fund would have given results in line with hypothesis?
Or
Is it entirely wrong to do this comparison at all?
I am not sure if I know answers to these questions.
What do you think? Do let me know of your thoughts about the idea of splitting money (MF & RD) as done in first scenario and also about my assumptions regarding trigger points.
Did you include the interest received on the funds invested in the market crash fund ?
Dude this is so misleading. You can't ignore interest. Please rewrite this post with that calculation. I urge other readers to not be misled by such half baked posts which present only one side of the story
Dev,
Firstly, definitely interest should be factored in the calculation because we are considering very long periods of RD 96-00, 02-08, 9-11 and 12-14. Even at a conservative 9% (it use to be higher in the late 90s), that would mean that the money would have compounded by 1.4 times for every 4 year period.
Secondly, there are two big variable assumptions here – how much to split and the trigger. If you know of Amibroker software then this is what technical analysts do – optimization and walk-forward – for example what combination of moving averages to use and what to consider as the input of MACD to achieve maximum returns. However, most of the times, these models return results which are a best curve fit – they look good on paper but in reality things are different.
I personally believe that people should, based on their age keep a proportion of their money in cash and multiple it using FD/RD and the other should constantly keep going to MFs and a third part to picking stocks if the person knows what is he doing. For a 30 year old like me, I'd do a 10-40-50 split. I don't have the data but I feel that money kept in MFs as SIP for 20-30 year period should do very well because some of the best money managers know where to park the money during bad days so that they outshine during the good days.
Dev , Really appreciate the hard work behind this analysis. i think this might have taken you to at least 2 weeks to write this blog.
Most of us are busy in our day job , so it is always wise to run sip and make a 2 year emergency fund and when market will crash put one year fund as lumpsum… I know if you calculate RD interest part it will definitely beat the return.. But truly how many of us can do this ??? Even i am a great fan of all your blogs.. i write my blogs too but here you gave a very complex analysis … Believe me to understand this blog calculation .. it took me 30 minutes …. then think about normal lay man…
When we are saying market is supreme.Complex calculations are part of high paid financial guy's job…as a normal person i generally refrain myself from this type of complex calculations.
My strategy is as below:
one mid and/or small cap/multicap fund(SIP)
One large cap fund (SIP)
One Balanced fund (SIP)
80C ELSS fund(max 2- i do this in lump sum)
Yearly if i find good stock then at max 2 addition to portfolio else not required.
In market crash add companies from one year emergency fund in 3 lot ,which i get at a cheaper price and good growth history and potential and business which i can understand .
that's it.
Every 6 quarter once performance check of all this fund…
This thought is entirely mine .. I may be wrong or right …
http://smartfinanceplanning.blogspot.in/
Hi Dev,
Here are few points that i can think off
1. To start with, having a MF (and that too one tracking top 200 stocks) may introduce too many variables (basically active management comes into picture).
Instead, you might want to do same analysis with a pure index fund or even better one stock.
2. Instead of NAV, you might want to use the PE ratio. See what Pattu calls as SI-PE (http://freefincal.com/are-mutual-fund-sips-suitable-for-disciplined-long-term-investors/)
Btw, once you have figured out a way when it does match the theory, it would be interesting to compare the strategy against, just doing a constant SIP in a equity oriented hybrid fund (HDFC Prudence or Balanced).
Thanks
This is excellent analysis. Two questions: What was the assumption behind RD interest, and what happens to the tax on RD returns?
While I agree that SIP would be the best approach for a period of 20 years, the problem is you don't know if you can really leave the money alone for 20 years. There are several milestones that could disrupt the process. So I don't do a SIP in an over-valued market (PE of 21+).
My plan is to systematically accumulate cash (via arbitrage funds since LT returns are tax-free) and balanced funds when market is hot (PE 21+), invest in Equity MFs below PE 21 and pump in cash when valuations are really down (18 and below). How does it sound?
Dev,
While we are on the subject, do you have any thoughts on where should a person park their SIP – a regular MF or ELSS assuming that there is still some pocket left in the 1.5L annual tax allowance that the government allows.
No Manoj…
For simplicity, I ignored the interests accrued in Market Crash Fund…
But I guess a lot of readers think that it should have been considered too… 🙂
Hi Shan
I understand that ignoring the interest can have a 'impact' in overall results. But as far as I looked at the size of the fund being accumulated in market crash fund at regular intervals, the interest would not have had a game changing impact.
Its possible that it might help the MF+RD win this particular battle, but I am not sure whether including the interests will have a huge impact, which I actually expected when I started with my analysis. But I will still analyse both scenarios again to see the impact of including interest component.
Hi Shan
I understand that ignoring the interest can have a moderately big 'impact' in overall results. I just chose to ignore it at first because I thought that RD+MF combination would win the battle without the need for considering the impact of interests. But it seems that results show that a little help from interest component is required to get the whole picture. I will analyse both scenarios again to see the impact of including interest component and do a follow up post soon.
Saurabh, I would suggest to go for good ELSS (as your 1.5L is still left).
Hi Saurabh
I will do a follow up post to see how a similar approach (but including interests) fares over the said period. My reason for ignoring it at first was the thought (or rather assumption) that RD+MF combination would win the battle without the need for considering the interests. But I guess my assumption was wrong 🙂 Thing will get clearer when I do further analysis.
Secondly, this is a just a theoretical exercise and I very well know that in reality, returns for investors like us tend to be way off what simple mathematical calculations show. 🙂
But I like your thoughts on splitting money between cash, FD/RD & MF. My approach is quite similar, though weightage are slightly different. And as an extension to your thought, it points to the importance of asset allocation. At times we are so concerned about picking the best MF/Stock that we forget that its really the concept of asset allocation, which has a bigger say in long term
wealth creation.
Thanks for sharing your thoughts.
BTW Shan
I have highlighted the statement in the post about ignoring the interest part. So that nobody misses it. 🙂
Hi Abinash
Thanks for sharing your approach. I really like the simplicity of it and frankly speaking, at times (or rather at most times), simplicity wins. So I am sure your approach will do wonders 🙂
My intention of doing this analysis was to see how we, as average investors can make use of concept – ‘Invest At Lows’ and do it practically. I understand that being in regular jobs (me included), it’s tough to keep abreast with NAVs and track when they are trading at X% discount to
their 3 year highs. But it’s just a trigger point which I thought made sense. It can be any other point too. And that is why readers should take this analysis as starting point to see that it makes sense to invest at bottoms. And be prepared to invest at bottoms by making a personal market crash fund.
But I agree with you that this is not a very simple way of proving something 🙂
Thanks
I guess any investment in market linked instruments should be done on the assumption that we dont need that fund for next 5-10 years. And probably then comes tax savings part of that investment. If money is not needed for few years, ELSS seems like a good bet.
I second you 🙂
Dev – First of all thank you for this post.
This is a very interesting idea and something I've personally pondered for quite some time.
I know a lot of readers have mentioned RD interest component but that's beside the point that you are trying to make and my guess is even if included won't amount to a whole lot which was the expectation to begin with.
My two cents on questions you asked :-
1) No point of taking an index since our preference as investors is always over actively managed funds instead of index funds. HDFC 200 should do the trick to test the hypothesis since it's a tried & tested active fund.
2) The split too sounds good to begin with. I guess you could try a 75-25 (SIP-RD) as well after you explore point 3 below and it doesn't work! 😛
3) My only additional comment is wrt to deployment of lumpsum RD. I believe trigger should be definitely more than 20%. Ideally the lump sum should be deployed just once/twice every 6-8 years when there's literally a bloodbath. Don't have much market history to tell how deep that cut should be. But, my guess is trying it out with 3 scenarios : 30%, 40% (My preference) & 50% would be good.
Hi Anand
1) Choice of HDFC Top 200 was because its a decent, respectable funbd which has survived the market cycles and delivered better returns than overall markets (index). But an index fund based
analysis can also be done here.
2) And frankly I did give a thought to using PE Ratios (of index) as trigger point as I myself regularly track it to gauge investor sentiments 🙂 But since most readers might not be interested in finding index PE and would rather find it easier to simply check NAVs, I thought it made sense to choose a NAV-based trigger.
And choosing a balanced fund and repeating the analysis might give some further insights. By the way, thanks for sharing the link to Prof's amazingly detailed analysis. 🙂 He is really the Excel-Man of India 🙂 [#Respect]
Hi Govar
I have ignored the impact of interests in this analysis. And hence the tax part of it too.
But the approach which you suggest seems good to me. Coincidentally, I analysed similar approaches in 2012 and it made sense to me. I did two articles on these approach which I would like to share with you here:
(First)
http://stableinvestor.com/2012/07/do-indian-stock-markets-bounce-off-pe.html
(Second)
http://stableinvestor.com/2012/08/buy-low-sell-high-how-to-use-markets-pe.html
Let me know what you think of these.
Appreciate your feedback Jay.
I would really be interested in doing the scenario analysis required in point 2 and 3 (and ofcourse many more which other readers have requested). But not sure if I will be able to do all of them as I have a finite bandwidth here ;p
But yes…will be doing a follow up post soon to cover major aspects.
Thanks.
Of course Dev. In that case I would just suggest point 3 with the trigger at 40%. That would be my only request. Thanks!
Hi Dev,
It would be interesting to know that – for the period under consideration, whenever the '>20% discount to max NAV in last 3 years' trigger has been activated, how deep market has fallen and for how long it stayed below this mark. The longer and deeper market stayed below your trigger point, pure SIP would have scored better than RD+SIP. Sorry, but I'm really not able to comprehend your second graph, so couldn't read much in it.
Hi Dev,
1] It would be interesting to know what was wt avg NAV of MF investment under the two strategy during two bear markets of 2000-2003 and 2007-2009 and the peak NAV in subsequent bull period. May be we'll find some answers there.
2] Also by design of strategy, there will always be less capital participating in equity during bull phases of market under RD+SIP as compared to SIP.
3] Instead of having a single trigger to dump all RD monies, we can have sequential triggers for step wise deployment. But it may make things complicated, so why not stick with pure SIP 🙂
Hi, The fund is taken only for the purpose of concept and data comparison. It is a fund of fund with disadvantage of tax. It is not an investment idea. The returns shown will further come down if we include tax.
Hi Krishna,
Have you tested the same with different numbers? I mean, the following combinations:
1. Investing crash funds when monthly average NAV is 10% less than highest 3 year NAV
2. Investing crash funds when monthly average NAV is 20% less than highest 5 year NAV
3. Investing crash funds when monthly average NAV is 10% less than highest 5 year NAV
and many more you can think of……
Have you tested the same with different numbers? I mean, the following combinations:
1. Investing crash funds when monthly average NAV is 10% less than highest 3 year NAV
2. Investing crash funds when monthly average NAV is 20% less than highest 5 year NAV
3. Investing crash funds when monthly average NAV is 10% less than highest 5 year NAV
and many more you can think of……
Fund Manager Matters….
Hi Dev,
On point 1) – yes, checking this hypothesis against the index fund would provide us the starting point (or base line numbers to consider for this exercise, so to speak). The deviation in returns of HDFC200 can then be attributed to the fund management part of it.
On point 2) – i meant taking the PE ratio of the fund as the trigger point.
On the balanced fund – my point was – once you have found some combination (of percentages and trigger points with RD and HDFC200) that matches the hypothesis, then to take that approach and compare it with – just doing a SIP (in this example of Rs 10,000) in one of HDFC Prudence/Balanced (or one of the top balanced fund around from that time). The intent is to find if we get better or worse results (my sense is we might get equal or better results – fundamentally reflecting some form of automatic 'buy low sell high') and in which case (and given this time horizon of 15 years), the solution is simply good balanced funds!!
Thanks
Got it Anand..Let me check it out…
Hi Abhijit
I haven't tested all the above mentioned scenarios but I feel that 10% ones might not be much useful. And that is because occurrences of such scenarios (NAV 10% below 3Y or 5Y highs) might not be that rare and would be happening regularly. But checking out 20% ones might be more useful…
Guess comment above got duplicated…
Answering again:
I haven't tested all the above mentioned scenarios but I feel that 10% ones might not be much useful. And that is because occurrences of such scenarios (NAV 10% below 3Y or 5Y highs) might not be that rare and would be happening regularly. But checking out 20% ones might be more useful…
Yes Karthikraja… but I guess that when comparing a MF+RD and pure MF, we are anyways using the same funds and hence the same fund manager…but ofcourse weightages are different.
Dear Dev Ashish,
I am very much impressed with your great in depth work. Many peoples have commented on this system and you have also choose some of them and analyse again to give optimum yield on investment. I think one more system you should calculate which is yearly portfolio balancing to give highest possible yield.
There are some reason for it to give high return. One of the most benefited result is that
1. In yearly portfolio balancing, if nifty P/E is going higher and equity share going up and doing balancing, it is one type of profit booking and reinvested in secure return generated debt. AND
If nifty P/E goes low and debt share is going up and doing re balancing, it is one type of low cost investment in equity at lower price to gain more in future when nifty goes up.
So you are requested to do same analysis by investing 5000 in equity (MF) and 5000 in debt (RD, please include interest too). Every year re-balance portfolio to 50:50, if its changed some what like 54:46 or 45:55.
You are requested to share me this calculation please.