Powered By Blogger

About Me

Retired chief investment officer and former NYSE firm partner with 50 plus years experience in field as analyst / economist, portfolio manager / trader, and CIO who has superb track record with multi $billion equities and fixed income portfolios. Advanced degrees, CFA. Having done much professional writing as a young guy, I now have a cryptic style. 40 years down on and around The Street confirms: CAVEAT EMPTOR IN SPADES !!!

Thursday, December 09, 2010

Treasury Bond Strategy Factors

Today's post builds upon yesterday's entry on the Long Treasury bond. Here I want to look at
strategy factors and the difficulties involved in devising workable strategies.

Long term interest rates are sensitive to the trend and level of short term rates. There are
effective models one can use to get a fair bead on where bond yields should be when short
term rates are at moderate levels. Such is not the case when short rates are at extreme levels.
There is an extreme now with the 91 day T-bill rate at just around 0.15%. Moreover, since the
Fed has no intention to raise short rates in the near term, there is not a solid model application
here for the bond. My very long term model for the level of short rates based on inflation
factors implies the T-bill should now be yielding about 3.0%. My very long term model
for deriving the long bond yield from the short rate says I should multiply the bill rate by 1.5x.
The model implies that the long Treas. should now be yielding 4.5% (3.0 x 1.5), which is
close to the present yield for the bond and which also suggests bond players are assuming
that short rates will eventually rise moderately. This hypothetical run-through is interesting
nonetheless.

To protect purchasing power, a bond needs to provide current return and re-investment of
interest received return  which exceeds inflation by a meaningful degree. With the current
CPI running about 1.2% yr/yr, the old rule of thumb is to add 300 basis points to the CPI
reading to get get a fair yield for the T-bond. This informal model puts the "proper"
yield for the bond at 4.2%.

Another approach I have used in recent years is to deduct a 3% inflation assumption from
the yield on the 30 yr. T-bond.  Experience shows the bond tends to rally in price when
there is nearly a 200 basis point premium over the 3% CPI assumption and to not do so
well when the premium is only 100 basis points or less. See chart. (Note too, the
sensitivity in yield to industrial commodities  prices such GS's industrial metals composite).

There have been few periods in US history when the inflation rate has sustained above 5%
for an appreciable period. Mostly, these inflation surges have come around war time when
resources are heavily in demand. However, if you wanted to look out past a few years and
were concerned that inflation could average 3% for a sustained period, then my work
suggests the T-bond yield would have to rise to 6% before it offered decent value. And, it
will do precisely that on evidence of a sustainable acceleration of inflation pressure from
the present low level.

With a rise in the Treasury's funding requirements, bondholders should expect a premium
to be built into the long bond yield to cover a much heavier supply of new debt and a
higher level of re-funding. I do not see that yet, but if confidence grows further in other
riskier markets, it may appear and could add up to 100 basis points to the bond yield.

In summary, the T-bond is reasonably valued now given the low levels of short rates and
inflation. Obviously if the recovery continues to advance, broaden out more and solidify,
then it would be no stretch to the see the long Treas. move up to 5.25 - 5.50%.

No comments: