A Simple Strategy that beats the Market — An Introduction to Quantitative Finance, Part II
Note: This is last part of our series on Basic Investing, Python and Quantitative Algorithms from our first Quant meetup in Raleigh-Durham. So its a summarized version of our actual presentation. Read part I here.
“More money has been lost reaching for a little extra return or yield than has been lost to speculating. It is one of the greatest temptations new investors face when building a portfolio” — Benjamin Graham
Quant investing
After our introduction to basic investing principles and asset classes , now we move on to quantitative side of investing.
My simple definition of Quant strategies :
Intersection of Statistics, Data Analysis and Finance to generate returns
I use the following framework to come up with a strategy.
GSM — Goal, Signal and Metric
- Goal — Your expectation — Low volatility? Or capital protection? Or just crazy returns ignoring risk. Understanding your strengths and weakness is key. Also understanding risk is very essential.We will talk about this in a future meetup.
- Signal — This is usually the crux of your strategy, I typically start with some of the old school investing techniques — Value Investing (e.g. Dividend Yield), Momentum strategies( e.g Moving Average) .
- Metric to capture that signal — Example — 1 year returns or Simple Moving Average > price, price/dividend > 3 (or some other number)
Limitations of Buy and Hold
- Draw-downs & Emotional stress during market crash — Drawdowns are how much market has crashed from its previous peak value. Look at the graph below, these are the drawdowns of S&P 500 in last 20 years
When the market crashed 45% in 2008, you are likely to panic and sell than hold as you think. This is despite the knowledge in long term stock markets recover. People are unable to control their urges (you cant blame us that is our entire life savings or retirement money).
- When to SELL? — Similar to Point above, except people always want to avoid the big crashes as they are hoping to retire sooner or pay for their expenses like buying a home. They would love to protect their capital ,at-least some percentage.
- Tail Risk — This one is straight out of a statistical text book , look below for following chart, Buy and Hold in 2 different 10 year periods and look at the difference in returns.
So tail risk is basically returns of a 10 year period (so much for long term investing!) can vary widely — 5.77% Vs 15.08%. To put it in Dollar terms, a 10,000$ investment would have given you 17,532$ and 36,238$ respectively.
That is like a difference of 18,706$!, The difference is more than returns generated in the first period!.
Advantage of Buy & Hold : Its very SIMPLE, Straightforward.
“Investing is Simple but NOT Easy “
— Warren Buffet
To prove the above quote , Just take this following stat!
Average Investor return: 3.64% vs SPY: 6:94% (Last10 years).
What Strategies you will hear from us at pfQuant ?
“Investing is simple but not easy” — we try to follow this.
- No fancy algorithms — No neural nets, no deep learning. We believe those complicate things and we cannot easily diagnose the flaws.
- Just play on Behavioral / fundamental aspects of investing.
- Strategies that can apply to all asset classes with minimal changes.
Most Important, Build systems to stick to our strategy — It could be as simple as setting a calendar reminders to open your excel sheet!.
Without much ado, we move on to our first strategy.
MOMENTUM INVESTING
“The premier market anomaly is momentum. Stocks with low returns over the past year tend to have low returns for the next few months, and stocks with high past returns tend to have high future returns.” - Eugene Fama
Guys at ThinkFound have a better explanation than I can possibly write.
Momentum is a system of investing that buys and sells based upon recent returns. Momentum investors buy outperforming securities and avoid — or sell short — underperforming ones.
The notion is closely tied to physics. In physics, momentum is the product of the mass and velocity of an object. For example, a heavy truck moving at a high speed has large momentum. To stop the truck, we must apply either a large or a prolonged force against it.
Momentum investors apply a similar notion. They assume outperforming securities will continue to outperform in absence of significant headwinds.
Why Momentum works?
There are many theories, but important ones:
•Behavioral aspect: Rising prices tend to attract buyers while falling prices tend to attract sellers.
•Over reaction and herd mentality.
This strategy is also called Trend Following and has been back-tested over 200 years across asset classes, This edge has remained whereas all other factors including “Value” have disappeared in recent years.
ALGORITHM — Simple Strategy that beats the market
Risky Asset : More volatile asset — here we use S&P 500. SPY is the Exchange Traded Fund that tracks S&P 500.(Trivia: Its the oldest ETF!)
Safe Asset: Less Volatile asset, Where Capital is preserved — CASH (or Treasury Bonds — SHY, TLT)
Do the following every Month (first day of a month is a good habit)
If Current Price > Price 12 months ago
Buy Risky Asset — S&P 500 index(SPY here)
ELSE
Sell SPY and Go Safe Asset (here CASH)
That is all very simple — How did this model actually perform?
A 10,000$ investment, would have returned $34,733 compared to $32,162 Buying and Holding SPY in the same time period. Better than this is the sweet spot = In 2008 sub-prime crisis this strategy had a draw-down of -16.23% compared to 50.97% of SPY. Not bad for such a simple strategy!.
In our last post, we showed only 1 out of 20 fund managers outperform SPY. This one just did it ;) .
Limitations of this strategy
- If Market crash is the risk for Buy & Hold, Whipsaw Markets for Momentum. Markets that keep changing directions often can erode the gains.
- In general — Higher turn over, Higher trading costs and Tax in-efficient compared to Buy and Hold.
- It doesn’t outperform year in and year out.
Also as more people are aware of this strategy, there are many implementations of the same (esp Post 2009). So this edge may not continue to exist in future. As always
PAST PERFORMANCE IS NOT INDICATIVE OF FUTURE PERFORMANCE.
All this is great, but I don’t have time for that or knowledge in XYZ and all other things you can say —
There is an ETF for that! — These are different implementations of Momentum strategy. Some of my favorite options are:
In our next post, we try to show metrics that matter and we will be dissecting the above strategy with numbers and see if this actually qualifies as a good strategy against those tests.
Posted by Prashanth Rajendran for pfQuant. We are a quantitative meetup group hosted in Raleigh-Durham.Details here
