Financial advisory can be divided into components. One which is generic advice, the kind available at the cost of only electricity and internet usage charges in blogs, forums and elsewhere. Then there is personal advice which can be obtained for a fee from a planner or free from your LIC agent uncle. Unfortunately most confuse one component with the other, but let’s not open that Pandora’s Box today. Personal advice is personal and not much can be written about that. Generic advice on the other hand, is and must be (in my unimportant opinion) grounded in mathematics. That is the principles must be supported by *consistent* mathematics. All very well, what has this got to do with the post’s title?

FundsIndia, a Chennai based mutual fund distributor recently published a report titled, “Value-averaging Investment Plan with Indian equity funds – An analysis”, authored by Srikanth Meenakshi and Shankar Bhatt. In this report they claim that the value-averaging method (VIP) of making periodic investments in equity mutual funds has a higher compounded annual growth rate (CAGR) (~ 1.66%) than the systematic method (SIP) for a chosen set of representative conditions. The authors have approached the problem with an academic mindset and have been good enough to share the details of their results with no hidden formulae: FundIndia VIP Backtest (Can be downloaded as an Excel file).

Before we proceed, some quick definitions for those unfamiliar with SIP/VIP/CAGR

SIP: Investor buys units of a mutual fund for the same amount each month.

VIP: Investor buys more MF units when the market (therefore the fund and therefore the investor’s portfolio) performs lower than *expectation* and vice versa. You can read more about this here. So you need to set an expected return, a min, nominal and maximum amount for monthly investment. The expected portfolio value is calculated with the nominal investment amount. You can read more about this in the report itself.

CAGR: Returns from an equity investment fluctuate every day. So to represent growth (or lack thereof) of the investment an *average return *is calculated. This average is not the usual average that we are used to (called arithmetic average) but a *geometric average* (This is because compounding represent a geometric progression).

Now getting back to the report, the more I stared at it (don’t ask why!) the more I became uncomfortable with the method used to arrive at the findings. The central result of the report is the uniform higher CAGR observed for VIP investments compared to SIP investments. To be brief the CAGR is obtained from the arithmetic average of the monthly difference in returns between VIP and SIP investments. Trouble is this method does not seem to reproduce standard results. For example if I have two recurring deposits, one with annual interest 10% and the other with 15%, the CAGR difference between the two is 5%. Whatever method I use to obtain CAGR for equity investments it should give me 5% for the RDs. The approach used in the report does not give me 5%.

So I wanted to see what others do to compute CAGR for VIP investments. Many use a standard feature available in Excel known as IRR or XIRR. Here is a terrific resource (You can read more about IRR/XIRR here). IRR/XIRR returns 5% for the RD example therefore consistency is not a problem. The point of the opening para is: If we wish to claim VIP is superior to SIP as generic financial advice then we must use consistent tools to do so. IRR/XIRR is a readily available tool to do this.

Using the NAV data used in the report I have calculated the returns from VIP and SIP investments using Excels IRR function. The Excel file can be found here. I discuss some results below.

The report claims a uniformly higher CAGR for the VIP-mode. In some cases I find the CAGR difference to be much lower than claimed, in some case much higher and in some cases opposite (SIP outperforms VIP)! If we *must* choose between the two, then, I think the comparison must be made not just with returns.

First we should be clear about the purpose of investing: to attain financial goals. Thus the expected return, the nominal monthly investment and the final target portfolio should match our goal expectations (in part at least). This is important to evaluate which method is better.

The report is based on a study in which four mutual funds were considered for three 3-year periods (ie. 12 such studies) and the same funds were also studied over a 5-year period, thus a total of 16 such studies.

The compiled results can be downloaded here as a excel file.Here is a partial snapshot and some conclusions.

(a) 10 out of 16 times both SIP and VIP modes did better than the target portfolio value

(b) 5 out of 16 times neither SIP nor VIP did better. *Thus using a VIP does not guarantee success.*

(c) Only once did VIP alone better the target portfolio. *The point is if you can achieve your financial goal with VIP you are likely to do it with SIP as well!*

(d) Only 4 /16 times the total VIP investment is significantly lower than the total SIP investment (VIP inv. is only 63-73% of SIP inv.). This is an impressive victory for the VIP mode.

(e) 7/16 times total VIP inv. is marginally *higher* than SIP inv. *So no guarantee of this as well.* Three out of those 7 times in a losing cause (target not achieved)

(f) Coming to returns: 12/16 times VIP has higher CAGR than SIP. However 10 of those 12 times both VIP and SIP do better than the target portfolio. Only once does SIP fall short and once both do!

(g) 4/16 times VIP has lower CAGR than SIP and all of them in a losing cause (VIP and SIP fall short of target)

Let me stop here. Perhaps more can be said. Do let me know your observations.

In contrast the FundsIndia report claims that VIP outperforms SIP *all* 16 times! **Note: **My study is based on the NAV information given in their report. The total VIP/SIP amounts invested and final portfolio values are all the same in both cases. Only the target portfolio differs by a small amt (~ 1%) due to a minor difference in calculation and will in no way change the above conclusions.

**So what now?** First let us understand that the report was made with a singular aim: to promote FundsIndia’s value-added VIP service (it is a value addition no doubt there). If you are asking which is better wrt Indian equity funds then I would say it is not conclusive yet. Perhaps one should do this study for more number of years. I am working on a comparison with historic Sensex returns. Perhaps that will throw some light … perhaps.

The more important question is should I VIP or SIP? On the face of it there is nothing wrong with a VIP and it is certainly a good concept. However, using it depends on ones investing style. If I were to use a single index fund for investing then by all means I should do a VIP. In the long run I am likely to invest lesser than a SIP for at least the same kind of returns. This approach is not practical if use more than 2 MFs. For each MF I will have to specify a max. investment amount. For a VIP to work effectively the maximum investment amt. should be quite higher than the nominal amt. (wrt to which target portfolio value is determined). So I need to ensure regular cash flow. Obviously this gets tougher with more such schemes. Personally I would like to spread my eggs around a little bit. So I would prefer investing in about 4-5 schemes to limit my down-side risk (more important than returns). So a VIP is not suitable for me. Perhaps my style is less than optimal, perhaps it is not. I don’t care. I am comfortable with it and will like to stick to it. So you before you align yourself with their report or my results understand your personal situation and use a style which suits you best. Remember hindsight is always 20:20. Once you choose a suitable style you are never going to know the result of the other choice. Nor should you bother to find out. If you are new to mutual fund investing and have just started investing, VIP can be a good option. If you are yet to begin then you have no business reading about VIP! Get your KYC done and get started. You can worry about VIP later.

To conclude, I have nothing against FundsIndia or their report or against VIP. My point is if you want to make a judgement you should choose consistent techniques and fundamental yardsticks. Whatever the results, choices depend on personal circumstances. The key is to remain contended with those choices. An ‘*intelligent investor*’ is necessarily a ‘*contended investor*’.

Pattu,

Thanks for the detailed analysis. At first reading, it appears that both of us were trying to measure different things. We were trying to measure pure outperformance between the two methods, while you are measuring the methods’ ability to reach target. Also, we wanted a “bottom-line” number that would highlight value-averaging’s potential effectively. However, let me go through your analysis more closely before drawing conclusions.

Whether or not VIP is “suitable” for an investor is a subjective call. From an advisory point of view, I have often told people that they should get comfortable with SIP and monthly investing in equity markets before venturing into VIP or other FundsIndia value-adds (like Flexi SIP). The understanding required for a good VIP/VTP implementation is non-trivial, as your post clearly indicates.

Some of our investors use value-averaging very effectively in FundsIndia (especially VTP), and have continued to do so over 2-3 years now.

Thanks again for spending time on this analysis and writing up on this very interesting topic.

Regards,

Srikanth

PS: Our invitation for you to visit our offices has been open for long; Hope you’ll take it up soon 🙂

Dear Srikanth,

Thanks for commenting and taking this in the right spirit. Yes I have focused on the methods ability to reach a target. However I have also looked at the IRR returns. In fact in some cases the IRR return difference is higher than what you quote. In some cases much less and in some the other way round. I will wait for you to look at it in detail. I agree with your views on suitability of VIP.

Back testing in Investing is as useful as you know to torture data. You can first arrive at a conclusion, then arrive at a data. Best is to see how a ROLLING avg works over a 10 year period in RANDOMLY chosen funds. After all that is what will happen going forward will it not? More difficult to implement on your own…that is all….

Cherry picking data is one thing, cherry picking math is quite another. Even with consistent math, aftcasting (as some call back testing) has its own limitations as does forecasting with a dice (Monte Carlo). We do need them … for a foot hold. Trouble is, stare at them too long and the underlying assumptions fade.

Agree fully. KISS – Keep It Simple Silly must be the financial mantra.

1) For saving for financial goals people *must* invest some amount each month. Trying to vary that can yield disastrous results. Simplest VIP is to do a regular SIP (for the goal) and try to contribute another 2-3k or so on days whenever markets fall nearly 2% (say like Apr 4, 2013). Never miss SIP no matter what.

2) If the mandated savings amount is not consumed human mind will use every opportunity to squander it so assuming someone will increase investment next month or in future because they are short of the goal is a fallacious idea. I did not read the numbers in detail yet (Both FundsIndia and FFC) but I am quite sure that most financial models work on paper but fall short on implementation – almost always!

Hi Anand,

Fully agree and try to implement both 1) and 2).

Thanks

This is quite good and much detailed work. My observations:

1. Why should active funds be part of a mathematical analysis? Some of the underlying fund strategies may actually be “some form of Value-Investing” which is being compared and that will definitely influence the results, sometimes more sometimes less. But that will just increase the inherent error within the analysis. Yes, since this is a more detailed and better mathematical analysis of the same data as of Fundsindia, this part can be skipped. This should get corrected when you will use the Sensex calculator. Looking forward to that.

2. The more samples you have, as Subra mentioned, more will be the Confidence Interval of the observation. Rolling returns also randomizes the data more.

3. Can there be a more realistic analysis in terms of total amount. If someone has 5k per month and by VIP formula he has to invest only 4k a particular month, what is to be done with the leftover 1k? That will also generate some return (4% if left in a large bank’s savings account). What really is the practical utility of investing a less amount of money to get more CAGR but a lesser total portfolio value? I have seen that in some of the observations in these data. Also, when one is considering to invest for 10-15 years, the total amount of investment makes a lot more difference than just the CAGR, especially when the overall difference in VIP and SIP is not that much.

I hope these things will be better addressed in the longer Sensex calculator.

Regards

Ramesh

Hi Ramesh,

Thanks. Excellent points as usual.

1. Active funds: Yes I noticed this in some of the 3 yr periods esp. the middle one.

2. Yes the Sensex calculator should answer some issues. Its nearly done, trying to put in an option to select no of years in addition to the full duration (1980-2010 or so)

3. Terrific point. VIP invests more than SIP in a couple of runs and still ends up short. In some others VIP has a significantly lower sum invested. In most the difference is about 10% only and in most of those runs both bettered the target. Longer duration testing might give VIP the edge.

Just one more point, since this is a mathematical analysis. If you can add one more thing, in which instead of the VIP method of Increasing the investment amount during the bear month, you Decrease the investment amount (the Momentum SIP thing). What results do you get then? Mathematically, it will give you a corollary statement. I do not know how you will do that, but just flipping the addition factor of VIP should be able to do that.

In general, I believe that since over last 20-25 years, equities have provided better returns vis-a-vis cash, the SIP method (in equities as fast as possible) should generally tend to be more beneficial than VIP (part equities and part cash) method.

Yes will try that.

Reg. SIP vs VIP over the long run. You are right! I have sent you the results. Intuition is always more valuable than brute force math.

Thanks.

Dear pattu, once again a healthy food for our brain. Thanks for such nice cooking. coming to the discussion, at the end of the day, as an Investor I’m able to get my goal or not is the most important thing. No matter what vehicle is used be it plain SIP, Flexi SIP, STP, VIP, VTP & so many cousins of all these, end result is important to me & interestingly that’s the most interesting part as neither Mathematics nor Science or even Art can guarantee that. In dear Subra’s word ‘Pure Luck’ may do it for us. 🙂

Thanks Ashal. Very true.

I suggest you read this excellent primer on this subject titled Value Averaging: The Safe and Easy Strategy for Higher Investment Returns by Michael E. Edleson. The book provides statistical evidence suggesting that Value Averaging produces 1% additional returns over Dollar Cost Averaging (or SIP in desi lingo). The author provides a step by step introduction to the subject supported by exhaustive formulae, simulation and excel based examples.

Thanks, but I am not too interested in US-centric results when I see no evidence of it here.