Tag: quant

Market-cap Classification and Illiquidity

Twice a year, AMFI is tasked with categorizing the universe of stocks into “large/mid/small” and funds with specific market-cap mandates are allowed to invest in only the corresponding set of stocks (amfiindia, sebi). We have a report that shows these changes over time (stockviz).

Given the massive flows involved, what is the prospective impact on liquidity as stocks get promoted and demoted between these classes?

We measure illiquidity using Amihud’s (Illiquidity and stock returns, 2002) illiquidity measure ILLIQ.

Between market-cap classes, the differences in liquidity is large enough to warrant a log-scale:

With this large disparity in mind, we can now look at the impact of migrations.

Large-cap Exits

Mid-cap Promotions

Mid-cap Demotions

Small-cap Promotions

It appears that stocks that get promoted from mid-caps to large-caps turn a bit illiquid. Otherwise, most migrations have negligible impact on their forward six-month illiquidity measure.

Code and charts on github.

Volatility Lookbacks

Volatility is calculated over a time period – the lookback. While developing a strategy, it is typical to try a range of lookbacks and pick one that looks reasonable for the strategy being built. However, is there an “ideal” lookback period?

This is where a volatility signature plot comes into the picture. It is typically used in high frequency trading but there is no reason not to use it on a lower frequency time series.

If you plot the distribution of volatility over different lookbacks, this is how it looks:

Ideally, you want the box to be small, the median in the middle and the wicks to be short. After all, if you are using volatility to drive a strategy, if the distribution of volatility itself is too wonky, then how do you trust the output?

Correlation Timing

The previous post discussed how high correlation environments have fat left-tails. Can correlation states be used for timing?

A quick look suggests that it might not be possible. Both LO (long-only) and LS (long-short) strategies that use the correlation state seem to underperform buy-and-hold.

There was some “crisis alpha” in using correlation for timing during the COVID crash of 2020. However, Buy & Hold ruled the roost both before…

… and after things normalized.

The slightly better drawdown performance of LO in some circumstances might be interesting for levered investors but the unlevered returns of these timing strategies are not much to write home about.

Correlation vs. Returns & Volatility

The previous post presented historical NIFTY 50 constituent pair-wise correlation distributions and discussed how high correlation environments persist. How do the 5th-quintile return and volatility look like?

If you focus on the 5, 10 and 20-day pair-wise correlations (T5, T10 & T20), you’ll see large left-tailed returns and high volatility in the 5th quintile compared to the others (1st, 2nd, 3rd & 4th).

Looking at this the other way, T10 – the 10-day pairwise correlation – has this profile:

It looks like if you dodge the 5th quintile here, you might be able to boost returns in linear strategies.