Thanks Abhishek for nice article. What should be the ideal frequency of running Algo. You can run it every minute or every hour or every day or every month. Any input ?
Is there a website from where we can gather the above charts or have you made the charts yourself? i just wish to know the source
Completely depends on the strategy you use. Typically you will design a strategy based on your preference of time period. For example, the time period I am most comfortable with is monthly and quarterly.
1 Like
Kotak Quant FundâŚ
Very detailed and thorough presentation.
They also shared Portfolio Construction Process & Kotak Quant Model and NSE Equity Index Values ( available in forms and downloads).
Kotak Quant Fund-Presentation.pdf (1.7 MB)
3 Likes
@basumallick Sir,
This is a great thread and sincere thanks for sharing the immense knowledge.
I am trying to develop a model for myself to transition to quant from discretionary stock picks. Will you be able to share your thoughts on following pls.
-
I prefer filtering the stocks only from Nifty 50/Midcap 50/Small cap 50 indexes (with 50,30 and 20% allocation) as I want to stay within index. Stock selection (Max 10) could be based on highest Alpha. While I need to do back test, wanted to know the opinion(s) on this approach pls?
-
You did mention the exit based on stop loss, is there a simple method to arrive SL in quant world? (Instead of considering fixed SL say 10%/20% etc.)
-
Letâs say if the stock is in F&O, is that a good approach selling the OTM calls to generate some additional returns? (Asking this since there could be a sizeable margin available and can do martingale if needed)
Thanks Much.
You can definitely select any universe of stocks as you wish or a combination of a few as you have mentioned. I have a friend who runs a fairly large capital pool of some private clients using only Nifty 50 as the universe.
There is no easy way of defining stop losses. The most common quantitative methods are as follows:
- Using ATR (takes care of volatility in the stock)
- Using a moving average (for example, stock closes below 20 or 50 day moving average)
- Fixed percentage (the easiest method)
- Using other indicators like RSI, MACD etc (like RSI going below 30 or MACD crossing below zero)
I havenât tried using F&O in my quant systems. Mainly because of my lack of knowledge and competence in F&O and also because I have found non-index options are mostly illiquid except for very large stocks. Maybe someday I will learn enough to try some experiments.
Nowadays, my experimentation is on commodities on MCX 
(I have to keep trying something new in life!!)
2 Likes
State of the Markets
Note: Why do we need to understand the state of the markets?
The state of the markets is like the âpitchâ in a test match. The captain needs to read the pitch and then decide on the playing eleven. For example, if it is a turning wicket, you may wish to play an additional spinner. Similarly, based on the state of the market, you might wish to change your allocations or stock selection.
One of the ways in which market health can be measured objectively is by monitoring the market breadth. It is purely data-based, unbiased and objective. Like any other indicator, it needs to be evaluated in sync with many other factors together and never in isolation. However, it is a good metric to keep track of. As of now market breadth is strong.
No of stocks hitting new 52 weeks highs has cooled down but it continues to be much higher than no of stocks hitting new 52-week lows (Universe Nifty 500).
Therefore the net new 52 highs and lows are continuing to remain strongly positive. (Universe Nifty 500).
% of Nifty 500 stocks above key long-term moving remains elevated.
Factor Investing - A basic introduction
As an investor, it is important to have some grasp of factor investing. If you hold a portfolio of stocks, the return it generates is due to something. This something is the reason for the return. The return generated could be because the stocks are paying handsome dividends or maybe because they are blue-chip companies. These reasons are known as factors.
We all do factor investing, either knowingly or unknowingly. Investors buy cheap stocks. That is the value factor. They chase momentum stocks, the momentum factor. Some investors prefer growth, the growth factor and some others may prefer smallcap stocks, the size factor.
Factors in use
Individual organisations, mainly large quant firms, have defined their own factors. For example, MSCI has defined the following factors in their process.
Factors can also be defined at a macroeconomic level with determinants such as GDP growth, inflation, interest rates etc being considered as defining ones.
Multifactor Models
In real life, we rarely invest if only one thing is going for us. We need more. In comes multifactor models. So, you might be interested in stocks which combine growth and quality together. Or momentum and liquidity. Or Value and Dividend Yield. Or low size, growth and momentum. The combinations can be many. But the idea is that we look for more than one factor to be in our favour when we search for stocks to buy.
NSE Nifty has rolled out strategic indices and has four defined factors - Quality, Value, Alpha and Low Volatility. It has created both single and multifactor strategic indices that you can look at.
The multifactor indices are:
- NIFTY Alpha 50 Index
- NIFTY Alpha Low Volatility 30
- NIFTY Alpha Quality Low Volatility 30
- NIFTY Alpha Quality Value Low Volatility 30
Use of Factors in Quant Investing
In India, most people equate quant investing with either high-frequency trading or momentum investing. It is neither. These are small subsets of the quant investing universe.
I personally prefer using those quant factors, which augment my own investing process. So, momentum, mean-reversion or multifactor models like quality+growth+momentum are where I try to focus on.
Some references if you wish to learn more on this topic:
https://www.blackrock.com/us/individual/investment-ideas/what-is-factor-investing
16 Likes
Some additional information and articles about factor investing in the thread " Factor investing & available passive instruments in india
1 Like
Hi, my question is on the re-balancing timeline for a Quant portfolio.
I have gone through a few quant/momentum based smallcases, and it seems that most monthly rebalancing ones have beaten the weekly rebalancing ones, on all timeframes. Especially, in a sideways market, the weekly ones have given 0 to insignificant returns b/w 2021 mid - 2023 mid.
Weekly: Pros
- Quickly captures a trend.
- Exits quickly on fundamental news, bad earnings etc.
Weekly Cons:
- Volatility, sideways movement. Bad returns.
- Transaction costs, hassle managing it on a weekly basis.
Monthly Pros:
- Reduces volatility. Many times stock doing well for 2-3 days donât do well for next few days and so on. But they will be in Momentum Radar for Monthly time frame. Ex: AngelOne
- Low transaction costs
Monthly Cons:
- Fails to capture major news or events, turnarounds. Ex: Momentum in Adani stocks post Election result.
Momentum in IT stocks in Mid of December due to positive on US exports. - Late into the trend, especially during Earnings season.
I already invest in individual stocks, and am considering a Quant based strategy as an alternative to Quant Mutual Fund holding. I definitely wonât be able to track earnings result and sectoral trend for 10-20 stocks which are part of my Quant portfolio. Regardless of that being a smallcase or picked on my own quant model.
1 Like
@Ascendant The period of rebalance essentially depends on the strategy. As you have already pointed out there are pros and cons in any period you choose.
One thing that people forget when they talk about quant or algos is that a strategy is an extension of the beliefs of the creator of the strategy. For example, I am, at the core, a long-term trend-following investor. I will always prefer strategies that are slightly on the longer timeframe unless I specifically want to exploit a shorter timeframe anomaly or opportunity.
This is why I typically prefer monthly or even quarterly timeframes. In my experience, the shorter the timeframe, the more statistical (technical-based) info becomes important. If you stretch your time horizon to months, quarters or even years (like the Magic Formula has a yearly rebalance), you move to the fundamental side more. The reason for this again is that there are more fundamental data points in longer timeframes.
6 Likes