S&P500 Fair Value Estimate - January 1, 2008
Document Actions
Gaining Insights From a Flawed Model
Pariveda’s quantitative S&P
500 fair market value model arrives at an interpretation of fair value based on
the aggregate operating earnings of the companies that make up the index over
the prior 1-year. While we don’t rely solely
on this model in making investment decisions it does provide insight on
aggregate value investors ascribe to the market and raises important questions
that we should address in understanding market psychology from time to time.
The model continues to signal a strong buy signal with fair value for the market as a whole at 2330, some 58% higher than where it currently trades. Our confidence in the model remains low for the reasons pointed out below.
The primary reason we don’t
ascribe more weight to the quant model in making our investment decisions is
that the data-set used to construct it, historical quarterly data going back to
1988, is really too short to provide statistically significant results. We can get index valuations going back before
this, but operating earnings weren’t reported, or re-constituted, prior to that
time.
Prior to 2004 the model did a good
job of estimating fair value and providing longer-term buy and sell
triggers. Since then however we’ve seen
a major divergence between our estimate of fair value and the market’s
willingness to pay up for this “value”.
It’s impossible to say exactly why
this deterioration has occurred, but it has happened as “value stocks” have
been out-performing growth by a significant margin – and within the 20 years of
data we’ve used, that simply hadn’t happened before. In our opinion, it’s likely that investors are
unwilling to value cyclicals, with the volatile nature of their earnings
streams, at similar multiples to more stable growth counter-parts. What this means is that the model is apt to
continue over-estimating fair value until growth stocks rotate back into favor
(which is now beginning).
It’s also interesting to note that
financials have made up a large and growing percentage of the index’s total
earnings over the past 4 years – but haven’t been rewarded for these
earnings. Look at any money center bank
one or two years back and they were being valued at 10-13 times earnings, a
much lower valuation than in times past.
Could it be that in aggregate the market was already pricing in, or
anticipating, that these earnings weren’t sustainable?
So, while we aren’t wedded to our quantitative model, we can and do gain insights by what it shows us on a quarterly basis.
When compared with 10-year
