|
PureBytes Links
Trading Reference Links
|
On Wed, 29 Oct 1997 19:51:48 -0600 (CST), you wrote:
>>Jim Roush wrote:
>>
>>> How many bear markets in this century have we had where there was not
>>> rising interest rates? None. Relax.
>
>This is an interesting statement, and one I've run across many times over
>the years. To verify its accuracy one must define two phrases: "bear
>markets" and "rising interest rates".
>
>If a bear market is defined as a 15% decline in the Dow or worse, and rising
>interest rates are defined as an increase in the 30 year bond yield or the
>three month treasury bill yield over any weekly period stretching from 1 to
>52 weeks, you can not prove the statement. When measuring such rate changes
>at the onset of every 15% or greater decline since 1951 (you don't have to
>examine the whole century for flaws to surface), you find that no
>measurement period of rising rates called every downturn. In other words,
>whether you measured rising T-bond or T-bill rates over 1 week or 52 weeks
>(or anywhere in between), every period exhibited some exceptions to the rule
>(i.e. rates were falling, yet the decline began).
>
>If you raise the bar to the level often used to identify a bear market (a
>20% or greater decline), again no measurement period for T-bonds survives
>without an exception. However, with T-bills 3 time periods did register
>perfect records.
>
>Before you "relax", however, you should be aware that all three of the
>"perfect" T-bill measurement periods were "rising" upon the commencement of
>the current downturn in the Dow.
>
>(Note: just because the "perfect" time periods were all positive doesn't
>mean that the Dow will fall 20% or more; that would require a separate
>study, as would the question of how many false signals of a decline "rising"
>interest rates signaled. All this study shows is that you can't "relax"
>just because interest rates may be considered falling by some measures.
>While it's clear that rising rates do often lead to a decline, they are not
>an absolute requirement.)
>
>Food for thought.
>
>Jerry Wagner
>Flexible Plan Investments, Ltd.
>Annuity Price Center www.annuityprices.com
> - the only source for downloadable daily variable annuity prices
>Member of the Society of Asset Allocators and Fund Timers, Inc, (SAAFTI)
> www.saafti.com
You are indeed correct that no one interest rate measurement can call
a market. However, I still maintain that a preponderance of evidence
in interest rate/ monetary evidence is best at calling long term
trends in equity markets.
"Encyclopedia of Technical Analysis" put a large number of indicators
(100-200) through a battery of statistical tests. Their conclusion
was that of the five best, one was monetary (a crossover 17 month of
the fed's free reserves, two used the change in the yield curve slope,
and two used moving average crossover of interest rates.
They were testing for predictability of 6-12 month moves.
James A. Roush
jar@xxxxxxxxxxx
|