Paratrade Systems Weekly Research News

22 Jul

Simplifying Factors That Drive Bond Returns

We have written about how we believe there are essentially only two things you need to know when you buy or trade bonds – inflation and economic growth. We have explained the inflation proxy (oil + the Auzzy dollar) and the economic proxy – the S&P 500.

Let’s now look at how these two factors affect returns individually and where we stand as of today in each case. We’ll start with the S&P. It is a reasonable assumption that if the economy is growing well then the S&P will be rising and that is bad for bonds. To study this we’ll use the ten day momentum of the S&P contract and look for at least a half a standard deviation move to filter out small inconsequential moves. The first question is: How do bonds perform when stocks are falling versus when they are rising?

Bond blog stock filter2

In case you can’t see the numbers, that’s a 56 point cumulative profit versus a half point loss over a very bullish data set (since July 2002).
As expected most of these gains come when the stock market was under severe stress but there was a lot of juice in bonds outside of those periods and you captured very little of them if you ignored the S&P and held bonds even when stock momentum was healthy.
So where do we stand today?

We are still in a bearish state since stock momentum is above the .5 standard deviation line but the spread is fading somewhat.
Now let’s look at performance when our inflation proxy is rising and falling. I shall define rising as when the 10 day momentum of oil*A$ is above one standard deviation and the reverse. We are raising the bar here to one full sigma since this series tends to be more volatile that the S&P and it is fair to assume that with so much else going on bonds need to hear a loud signal from inflation to be affected.

Bond Blog eflation mom

In falling inflation periods a bond holder earned 41 points while earning 12 when inflation was rising. This is valuable but it’s clearly a weaker variable than our economic filter.
Next we’ll look at combining the two.

Bond Blog combo filter

The results here are only slightly better than the economic filter by itself because the S&P tends to dominate the calculation. Let’s separate them so we can see if eFlation really helps. To do that we”ll look at setting the S&P momentum within a dull band while eFlation is outside it’s bands.
Let’s look at these two cases:

  1. The S&P >-.5 StdDeviations and eFlation < -1 Stddeviation
  2. The S&P < .5 StdDeviations and eFlation > 1 StdDeviation

Yes we are cheating a little since we want to limit the S&P momentum so it is generally confirming eFlation momentum because that is how we want to trade it. We don’t want to get into a meaningful position if the two things completely contradict one another. We just want to see how we do when the economic story is benign or helping and the eFlation story is speaking loudly.

Bond Blog combo3

Clearly this inflation proxy matters. There is only one period when it did not matter at all – April 2004. (Bonds fell significantly at the same time as stocks and eFlation fell.)

Bond Blog Current eflation

It is evident that my inflation proxy is falling and bonds have not been helped by it. We know why – stocks have been rallying like crazy. If we could say that the stock rally was tiring out then we would have a very good reason to load up.
Our S&P fuel gauge is indeed falling so bonds look like a good bet.

Add Comment