Alternative Viewpoints: “Liquidity Insurance”

CAPM / Alpha Theory 03 Mar 2008

In December, guest contributor Ranjan Bhaduri, CAIA examined the cost of liquidity by using a simple exercise that he called the “balls in the hat game”.  Bhaduri argued that illiquidity – a source of excess return – is often confused with “true alpha”.  Today Konstantin Danilov, CAIA, of Bank of America proposes a new type of security that could be used to hedge against the possibility of an illiquidity crisis. Danilov conducts buy-side manager research at BofA.  Prior to this, he was a trader at Cantella & Co.  He is also a member of the Program Subcommittee for Alternative Investments and Hedge Funds of the Boston Security Analysts Society.

His guest contribution below is the latest in a monthly series featuring members of the Chartered Alternative Investment Analyst (CAIA) Association.

“Liquidity Insurance”

Special to AllAboutAlpha.com by: Konstantin Danilov, CAIA, Bank of America

“Our current system of levered finance and its related structures may be critically flawed. Nothing within it allows for the hedging of liquidity risk, and that is the problem at the moment.   –William H. Gross, Chief Investment Officer of Pimco, New York Times Aug. 10th, 2007

Liquidity is a topic that is brought up often in the wake of a financial crisis. The crash of 1987, LTCM, Amaranth, and the current sub-prime crisis are all examples of the devastating impact of illiquidity.  Unfortunately, it is a factor that eludes the most risk management tools and risk/return models in modern financial theory.  For example, Value-at-Risk (VAR) and “portfolio insurance” largely ignored illiquidity (or assumed it away) and we were left with the consequences.

However, illiquidity in a less extreme form affects market participants on a daily basis in the form of everyday transaction costs. Returns on a stock position that exist on paper can quickly disappear when a manager attempts to sell it to capture a profit, especially if the stock is thinly traded.  I’d like to focus on these transaction costs in this posting.

I often think that this less extreme form of illiquidity could be better managed if we had a new financial instrument, similar in structure to that of a Credit Default Swap, that would enable market participants hedge this liquidity risk.  (For our purposes, I will define “liquidity” as the ease with which an operator can enter and exit it for a given block of securities.)

“A substantial part of unforeseen losses is due either to market jumps caused by illiquidity or to liquidation costs that substantially move the markets against one’s position.”

– Nassim Taleb, Author, Dynamic Hedging

Liquidity hinges on the asymmetry of information.  If a trader sees an unusually large sell order coming in, he will often hesitate before quoting a price at which he is prepared to buy.  He will often wonder whether the seller has information  about which he ought to know.  He may drop his bid (or completely remove it) until he can be sure there is no such information asymmetry.  As a result, the sell order might drive the stock price lower.  (The other option for the seller is to take a longer time to piecemeal the order, but that leaves him exposed to adverse stock price movement due to the arrival of new information.)

In a previous “Alternative Viewpoints” column on AllAboutAlpha, Ranjan Bhaduri proposed one method of measuring this “liquidity cost”.  But another way to put a price on it could be to construct a contract similar in structure to a Credit Default Swap (CDS), but based on the liquidity on a particular stock.

While such a metric wouldn’t measure default risk per se, default risk remains conceptually relevant to this idea.  When a CDS is traded, there is an asymmetry of information regarding the creditworthiness of the bond issuer.  As a result, one side is willing to insure the bondholder, for a premium, against default.

This type of structured insurance contract could act as a guide for markets to measure and price the liquidity risk of an individual stock over time.  There would be several factors involved in this hypothetical “liquidity insurance” contract:

Issuers of the contract

A CDS is issued by the banks that originated the bonds being insured.  Naturally, these banks are significantly more knowledgeable about the issuer’s credit risk.  Likewise, a “liquidity insurance” contract could be issued by the institutions that have knowledge of the true liquidity of a certain stock.  This could be an institution that is a) a dealer in a particular security or b) has better access to information about a trading in a particular security than the average market participant. Thus, a dealer that provides “dark pool” trading access, or trades actively in “dark liquidity” markets, would have better capabilities than a money manager with average trading capabilities.

Measurement and contract terms

There are several ways to measure liquidity.  One is to measure “slippage”, or the difference between the average execution price and the initial mid-point of the bid/ask spread.  This is a relative measure that can be used to compare different equity markets or different time frames for a single equity.  Other measures that could be used in combination could be bid/ask spread and Average Daily Volume.

For example, a contract could insure that a holder of 20,000 shares of XYZ stock (the contract notional amount), against a certain slippage amount when he tries to liquidate his position on the day of “exercise” (contract terms).  Note that the contract terms would be relative measures.  The average bid/offer mid-point might be a good metric to determine the slippage measure to settle the contract.  (This is one theoretical way this type of contract could be set up, and is just one example.)

Settlement and structure

The most difficult part of this theoretical exercise is to figure out how to physically structure the settlement of the contract.  One possible way is for the contract issuer (the insurer) to physically take the stock from the holder (the insured), and take on the liquidity risk by trading (i.e. selling or buying) it themselves.  The insurer would take the risk in exchange for the premiums received – assuming the position liquidity remained as expected.  This would have been taken into account when setting the initial contract term, much like the credit default probability calculations when a CDS contract is issued.  Under normal circumstances, the insurer would have no difficulty trading the stock within the stated minimum liquidity costs.  However, if an illiquidity event occurred, the insurer would bear the additional costs (just as the CDS issuer bears the costs of a default).

Obviously, I have ignored many technical details and oversimplified things for this brief posting.  However, the most important lesson to be learned from this exercise is that there can and should be a way to allow the market to price the perceived risk of a stock becoming illiquid. Furthermore, there needs to be a way to transfer that risk to those who are more qualified to bear it.  I believe this type of liquidity securitization would make the equity markets more efficient and allow participants to more effectively hedge their risks.

– K. Danilov, February 27, 2008

The opinions expressed in this guest posting are those of the author and not necessarily those of AllAboutAlpha.com, the CAIA Association or Bank of America.

Be Sociable, Share!

6 Comments

  1. Anon
    March 4, 2008 at 1:50 pm

    “Likewise, a ”liquidity insurance” contract could be issued by the institutions that have knowledge of the true liquidity of a certain stock.”

    You may be more informed than I am, but it’s my impression that the banks don’t have a very good understanding of liquidity — and therefore are not equipped to sell this contract.


  2. Ranjan Bhaduri
    March 4, 2008 at 2:42 pm

    Hi Konstantin,

    Thanks for the excellent post! This idea of a liquidity derivative similar to a CDS contract is something that can be explored further in the Winter 2007 issue of the Journal of Alternative Investments. Gunter Meissner, James You, and I wrote a paper that was published in that issue entitled “Hedging Liquidity Risk: Potential Solutions for Hedge Funds”. Without rapping too much, I think that this is an idea that can add tremendous value in portfolio management for Pensions, Endowments, and Fund of Hedge Funds. It is a simple, natural, and effective solution. Again, thanks for your post – it is indeed positive that people like you are thinking seriously about liquidity solutions.

    Kind Regards,
    Ranjan


  3. Ranjan Bhaduri
    March 4, 2008 at 4:17 pm

    As a footnote to my previous comment, I noticed that the Liquidity Derivatives paper that I cowrote with Gunter Meissner and James Youn is available on the internet for free(not for distribution):

    http://www.caia.org/ai/journalofalternativeinvestments/

    (it is the featured article on the CAIA website, at the right side of the page, and one may access it by pressing the link which states “Read the Full Text”).


  4. Vera Greenberg
    March 5, 2008 at 9:36 am

    Here are a few thoughts that I had about this article:

    – Given that securitization desks in banks and trading desks are separated by a “chinese wall,” how would the issuer of the bond have better knowledge of liquidity than the market?

    – How would the issuer of this instrument hedge himself? One way that I thought of, is with a put (this way when the issuer receives the stock, they have someone who is obligated to buy is from him). But figuring out the strike is problematic and buying a portfolio of puts is expensive. Any thoughts?

    – How would you know that the price moved because of the sale of this specific transaction and not some rebalancing at the end of the month by a pension fund? How would you model this — it would probably be using some proprietary data, as I do not know of any data tracking execution vs mid-price.

    This article definitely opens up the floor to a bigger discussion. Thanks!


  5. Gary
    March 7, 2008 at 7:09 pm

    You said it yourself, this contract would be worthless except in a “liquidity event”. In that case, what we really want to be managing is the ‘more extreme form’ (vs. the less extreme form that you talk about here).

    In a situation where we’re long a stock and want to manage this risk, we want something to protect us from downside jumps (which could also be caused simply by us exiting a large position in illiquid times, the less extreme form). I recall that LOR (of portfolio insurance infamy) offered something called “jump protection” as an extra product, I don’t recall the details of implementation unfortunately.

    I think this could be accomplished with an OTM put that is restruck daily (or weekly) at VWAP. On a $100 stock (r=5%, vol=15%), a 5% OTM put restruck weekly would cost $0.20/sh per year


Leave A Reply

← Kat to Jaeger: "Let's skip the nitpicking...how useful is modern finance theory, really?" Hedge Fund Portfolio Selection with Modified Expected Shortfall →