Suppose you invested in a stock market index only when it was "high" and compared the return obtained with that from an investment made only when the index was "low", how much do you think the returns would vary? How would these returns compare with an investment made regardless of index levels? Let us find out.
In real-time, market highs and lows cannot be determined. Therefore, it is common practice to compare the current value of the index with the average value of the last 100 days, 200 days, 365 days, 10 months etc. This is known as trend following. The average reduces the daily noise in the index values and provides a "trend'. Here are some examples: Nifty Valuation Charts Nov 2016: PE, PE, Div Yield, ROE, EPS Growth and a tool to calculate: Nifty Valuation Analyzer: PE, PE, Div Yield, ROE, EPS Growth Rate.
Therefore, for the purpose of this post, an index "high" will refer to a period when the current index value is higher than its 200 day average or ten-month average. Similarly, an index "low" is when the current value is lower than these averages.
I had recently calculated the total returns index of the S & P 500 :The magic of reinvested stock dividends!. Suppose we consider two monthly investments:
High-SIP: Made only during months when the index was high (greater than 10-month average). That is we check the index value on the first of each month. If it is greater than its 10-month average, we invest. Else we do don't.
Low-SIP: Made only during months when the index was lower than the10-month average.
Analysis with S & P 500
Suppose both these SIPs continued from 1st Jan 1900 to Sep 2016,
High-SIP XIRR (return) = 5.72%
Low-SIP XIRR = 5.85%
Perhaps a nice example to illustrate that in the long run, we are all dead! If you are worried about the low return, it is important to recognise that a Mutual Fund SIP is Hope, Not a Strategy! This includes "buying in dips" aka DIP-SIP.
But in this post, I am more interested in the difference in returns between High-SIP and Low-SIP. In this case, it is practically nothing. Might as well invest without worrying about market levels. I would have ended up with a return bet those two numbers.
Analysis over 15-year investments in S &P 500
Now, you might argue that this is not a realistic investment. Fair enough. So let us repeat this exercise for consecutive 15-year periods from 1900 to 2016. I will soon write a code to make this rolling.
Now, you can argue that buying-low always beats buying-high. That is indisputable. Also indisputable is the fact that the difference in return is quite small.
More importantly, no one is going to be buying high all the time. They would either buy-low or buy-anytime. Therefore the practical difference between trend-following SIP (as defined in this post) and a simple SIP would be even smaller.
How about the Nifty?
What about it?! The data set is too small to say anything one way or another. For what it is worth, here goes.
In this case, it was easier for me to consider a 200 day (daily) moving average for defining a high-sip and low-sip.
Before we go to a SIP let us consider five year lump sum returns. Suppose between 1999 and 2016, I invest in the Nifty only when it is low (value < 200 dma) and only when it high (value > 200 dma) and calculate the return after 5 years from the date of investment. In this each case investment is separate.
The difference between low-5Y return and high-5Y return (both lump sum) is 3% at both ends. If I had purchased regardless of market level, I would have a got a return spread in between these two.
The spread in return 45% to 2% for low (< 200 dma) and 42% to -1% for high (> 200 dma) is significant. Buying low alone is enough. One should also manage risk to reduce the spread. Risk reduction is not considered in this study, you can consult this, How to systematically reduce the risk associated with a SIP if interested.
Now over to the low-SIP vs high-SIP.
So between 20th July 1999 to 28th Nov 2016, if I have blindly set up a daily SIP, my XIRR would be 12.29%
If I had invested in days only when Nifty was above 200 dma (high-SIP), XIRR = 11.53%
If I had invested in days only when Nifty was below 200 dma (low-SIP), XIRR = 13.25%.
I have not adjusted the amount invested in all three cases to be the same as that would imply hindsight bias. Again, no one is going to set up a high-SIP. The difference between a normal SIP and a low-SIP is about 1% or so. This is over a 17.5 year period Whether this is important or not is up to you. As far as I am concerned, I have better things to do than to follow trends. I would rather continue a normal SIP and worry about managing the risk associated with my portfolio based on my financial goals. Also see: Do not enter equity markets if you do not know how to get out!
It is important to recognise that only investing has been considered here based on trends. One can also change the asset allocation based on PE or DMA. Studies have shown that such a strategies increase return obtained per unit risk taken, and not necessarily the actual return obtained. Read more: Is it possible to time the market?
Check out Part 2 of this study 🙂
Buying “low” vs Buying “systematically”: Surprise, Surprise!
Check out Reviews for You Can Be Rich Too
Ten reviews are available for our new book, You Can be Rich with Goal-based investing. Here is a sample:
The only book you need to get your hands on if you are lost in the financial jungle. Do it immediately. Simple, practical, crisp and precise.
The best gift that a young earner can receive.Kudos to the authors. - Karthik.
The book is also available from the Google Play Store (Rs. 244.30).