I would like to model liquidity effects in my risk model which is based on historical simulation. I would like to develop a practical solution that still captures liquidity effects. Most probably I have to treat equity markets and bonds markets separately but finally I would like to be able to apply some procedure for all assets in my multi-asset universe.

For stocks: historical simulation here is based on historical returns from market prices (taking into account capital changes and so on). What can I add here to either incorporate liquidity as an additional factor or to attribute parts of the return to liquidity risk.

For bonds: historical simulation here is based on zero-rate-curves and spreads mainly. What can I add here?

I am looking forward to an enlightening discussion. References and personal experiences are most welcome.


2 Answers 2


Autocorrelation of returns can be used as a proxy measure for liquidity of the asset.

The degree of serial correlation in an asset’s returns can be viewed as a proxy for the magnitude of the frictions, and illiquidity is one of the most common forms of such frictions.

A strongly liquid asset should reveal no serial autocorrelation.

You can perhaps build it into your model.


Couple points for your consideration:

  • At the time of order execution: You are most likely a liquidity taker and thus are rendered a service by those that provide liquidity and you compete for taking liquidity with other takers in the market. As such you need to have a firm grasp at the market impact of your order.
  • Liquidity can be extremely dynamic even intra-day and thus it is a major concern for anyone occupied with getting done optimal execution.
  • It is very important to understand the free float of a stock and how trading volume relates to such free float. A free float may suggest that a stock is highly liquid but if not many of its holders are willing to surrender the stock then traded volume may be extremely thin.
  • Even if trading volume is sufficient it may just mean that a stock is flipped without much accompanying price change. Think of it in this way: If there is a demand/supply equilibrium then the current price is a fair price and won't change much. So you may possibly see high trading volume but not much price change.
  • Liquidity in the fundamental sense really becomes important when a stock exhibits a high short sale ratio. Someone may place a tender offer for the outstanding shares and every last short seller will scramble to buy back the shares in order to return their borrows. An excellent example what happens at such times is the Volkswagen/Porsche story a while back. Just pull up a daily or weekly chart and you will see what I am talking about.
  • $\begingroup$ Thank you for your comments, @Freddy. Am I right that these apply for the stock market mainly? Do you have any ideas about the bond market? Furthermore - how can I capture this in a quantitative manner? I assume the big vendors of factor based models have some liquidity factor - but what can I do in pure historical simulation? Thanks. $\endgroup$
    – Richi Wa
    Commented Jan 28, 2013 at 8:16
  • $\begingroup$ except for the last point those also apply to cash bonds. It can even apply to treasury bonds where the cheapest to deliver affects liquidity in such ways that price is driven up and hence the bond is not the cheapest to deliver anymore. $\endgroup$
    – Matt Wolf
    Commented Jan 28, 2013 at 8:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.