Basis Risk - Future price not converging to spot

[hsuwang]

Hello David,
If basis risk exist for almost all futures contracts (at maturity), does that mean the futures price does not equal spot price at maturity? If this is the case, would it not create opportunity for arbitrage? I think I get the cause of basis risk (characteristics of futures contract differ from underlying position), but does this mean that if contract = commodity, there will be no basis risk? (ie. futures will converge to spot at maturity). Thanks!

 

[David Harper CFA FRM]

Hi Jack,

Hmmmm....I think basis risk has so much to teach. Yes, lack of convergence (futures to spot) clearly opens an arbitrage opportunity, but only because our model simplifies (excludes) frictions and technical factors. I really think you may enjoy this article @ econbrowser.
...long story short, Scott show why it's a convergence zone note a convergence point but more interestingly, how it failed in some truly commodity markets!

So i guess how i look at this is: if futures contract = *exactly* the commodity being hedged (and the terms under which the commodity is bought/sold, etc), then yes, our "clinical" cost of carry model implies no basis risk.

But, once again as is always the case, our model is not comprehensive. And mainly while it may incl fundamental factors (financing cost, storage cost, dividend, convenience), our model does *not* include technical factors like supply/demand for the futures contract. (a key theme: why do instruments often "trade rich or trade cheap" relative to the model? Often times b/c the model is not pricing technical factors).

So, in my view this statement is true for the cost of carry model: if contract = commodity, there will be no basis risk?
But not necessarily true in practice b/c the model cannot capture all dynamics

And the "basis risk," I would remind, refers not to the futures contract or to the spot (cash) market, but arises to the hedger, as a combination of those two things, who is using one instrument to hedge against the other thing. So, that's why i like to say "basis risk is always there for any hedge" .... but it varies, i guess at one of the spectrum (i.e., where the basis risk ought to be smallest) is a hedge against a financial commodity (no storage cost and specifications are very exact). But at the other, it gets really interesting when you look at something like the basis risk between a CDS and the underlying bond. I hope that's interesting - David

 

[David Harper CFA FRM]

...and you might find the arbitrage XLS i added interesting: http://www.bionicturtle.com/premium/spreadsheet/3.a._arbitrage/
i didn't use Hull's here, but rather Kolb's (unassigned) just b/c I wanted to model in transactions costs. And in this XLS if you add a transaction cost, the implied forward becomes an interval (lower and upper). So, it's not the same thing, but a similar logic could be employed, I think, to convert the cost of carry in Hull to reflect Scott Irwin's zone of convergence; i.e., a model that incorporates frictions like trading costs (or liquidity! as we speak i think a lot of models are being revised to add something that captures liquidity, easier said than done) and therefore doesn't show convergence to a single point - David

 

[hsuwang]

Thanks David for sharing this interesting article!

 

[narayananvenkat]

David..

Compared to Futures, basis risk is less in OTC derivatives. Can you explain this with an example.

venkat