Bid/ask spread reflecting endogenous liquidity risks


Hi David,

I'm working through the practice questions on Dowd Chapter 14, and on problem 7.4, why is it that answer C is true? I.e. how does the bid ask spread reflect endogenous liquidity risk? Is this because with some scenarios, like with level II quotes, you can not only see the bid ask spread, but also the size of the potential orders?


David Harper CFA FRM

David Harper CFA FRM
Staff member
Hi Lee,

Great point. I copied the associated reference from Dowd, see below. But please note: it's an example of myopia with respect to the AIM-against-text. I probably would not write it today, given my additional seasoning w.r.t. liquidity risk. I definitely would not include it in a mock exam (for example) because, to your point I think, it clouds a useful (working; aka, testable) distinction, that: in general, we treat the bid-ask as an (exogenous) external feature of the market. Further, in terms of modeling liquidity cost, Dowd used bid-ask spread only as an input (the risk factor) for exogenous liquidity risk. Conceptually, while both is probably true (and your statement may be true), I try not to write subtle questions that hinge on subtleties or, in this case, complex dynamics. So, I had in mind nothing other than try to quiz the below, but I'm not currently fond of implication that it muddies our "keep it simple" distinction: to model liquidity risk, use bid-ask for exogenous; use position size versus market (and elasticity) to model endogenous. Beyond that is super interesting but supra-exam. I hope that helps, thanks,

Dowd, emphasis mine:

Imperfect liquidity also implies that there is no such thing as ‘the’ going market price. Instead, there are two going market prices – an ask price, which is the price at which a trader sells, and a (lower) bid price, which is the price at which a trader buys. The ‘market’ price often quoted is just an average of the bid and ask prices, and this price is fictional because no one actually trades at this price. The difference between the bid and ask prices is a cost of liquidity, and in principle we should allow for this cost in estimating market risk measures.

The bid–ask spread also has an associated risk, because the spread itself is a random variable. This means there is some risk associated with the price we can obtain, even if the fictional mid-spread price is given. Other things being equal, if the spread rises, the costs of closing out our position will rise, so the risk that the spread will rise should be factored into our risk measures along with the usual ‘market’ price risk.

We should also take account of a further distinction. If our position is ‘small’ relative to the size of the market (e.g., because we are a very small player in a very large market), then our trading should have a negligible impact on the market price. In such circumstances we can regard the bid–ask spread as exogenous to us, and we can assume that the spread is determined by the market beyond our control. However, if our position is large relative to the market, our activities will have a noticeable effect on the market itself, and can affect both the ‘market’ price and the bid–ask spread. For example, if we suddenly unload a large position, we should expect the ‘market’ price to fall and the bid–ask spread to widen.

In these circumstances the ‘market’ price and the bid–ask spread are to some extent endogenous (i.e., responsive to our trading activities) and we should take account of how the market might react to us when estimating liquidity costs and risks. Other things again being equal, the bigger our trade, the bigger the impact we should expect it to have on market prices.

In sum, we are concerned with both liquidity costs and liquidity risks, and we need to take account of the difference between exogenous and endogenous liquidity. We now consider some of the approaches available to adjust our estimates of VaR to take account of these factors. -- Dowd page 310


@David Harper CFA FRM CIPM - I've seen you quote Dowd a few times (like above) regarding the bid-ask spread. I understand the concept but I've never seen it that the "bid" is the purchase price and the "ask" is the sale price. In every other literature I've read it's the opposite way. You'd sell at the lower bid price and buy at the higher ask price if you wanted to get rid of the security quickly.

Please let me know your thoughts. Thanks.


Well-Known Member
I think David said this in persective of the dealer that is he buys at bid price and sells at ask price. While if u look from the perspectibe of the investor he sells at bid price and buys at ask price. As a trade involves dealer and investor on opposite sides,saying either way is one and the same.