Back-Adjusted continuous futures contracts lead to NONSENSE

I have recently been wondering about claims that hedge funds can use futures contracts to increase the gains to be made on holding bonds. I have been looking at an ungeared version of a simple risk parity weighted portfolio which includes a number of US bond funds which concentrate on varying maturities and quality.

I pointed out yesterday that a simple trend following system can produce (in back- testing!) a very worthy and low volatility result on a portfolio of equal amounts of bond and equity funds in a 30 instrument portfolio, rotating monthly into the top 20 by momentum. Trend Following Without Terror

I then turned my mind to wondering how the likes of Bridgewater and AQR would gear such a portfolio. I am aware they gear the whole portfolio but I was primarily interested to see how they geared the fixed interest portion.

I had assumed they borrowed cheap relative to what they invested in. Perhaps they borrow short term and lend long term for the spread when the yield curve is positive. Perhaps the credit rating of their fund is good and they can borrow and invest in lesser credits for an uptick in yield due to lower quality.

Or can they use the futures markets? I decided to look at the ten year US treasury and see what a buy and hold of the futures contract would have achieved since 1985. I ignored the interest which would be received on cash not needed for margin. This is of course an important element in historic CTA returns but I wanted to see the return of the pure futures contract.

Which lead me once again to considering back adjusting methods and confirmed my distaste for the traditional panama style back adjusting method. At least in terms of assessing the profit to be achieved by buy and hold on futures contracts.  The effect of subtracting price differences on rolls vastly inflates/deflates the P&L.  You need to use raw contract prices and build an index or use some method which obviates the need to eradicate the roll gap in this distorting fashion. For instance you could achieve a more accurate picture by assuming you hold a spread of all maturities on each given day for the period you are trying to assess.

Here is a concrete example with the US Ten Year US Government Bond future (TY).

Panama style back adjusting gives and absurd CAGR of 14.4% since 2nd January 1985 to date.

Using a blend of maturities for each day throughout the back test gives the more or less correct CAGR of 4.65%

On a fully funded contract you would add to this the 30 day T Bond yield which would bring the CAGR up to a more realistic level, close to what you might expect on a constant maturity 10 year bond.

The combined figure (interest on T Bills and the return on the futures) seems to roughly equate to a constant maturity 10 year US government bond. So….good that you can gear into bonds using futures, bad that you will be completely mislead if you use back adjusted panama style contracts in your back testing.

Either create your own index or test with individual futures contracts on some roll scheme of your choice. Which must avoid delivery risk of course.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s