I.
Introduction
Investors often use financial ratios such as P/B ratio and P/E ratio to track the
variations in stock prices. Lee, Myers, and Swaminathan (1999) showed that a P/V
ratio could be better than P/B, P/E, and P/D ratios in tracking stock prices and
predicting the returns. The denominator of P/V is calculated using the intrinsic value
model developed by Ohlson (1995). Ohlson establishes the residual income
valuation method using the clean surplus concept that specifying the relations among
book values, future earnings, and dividends. His method allows one to estimate a
firm’s intrinsic value using contemporaneous accounting and information variables.
Specifically, a firm’s value can be expressed as its book value plus a linear function of
current abnormal earnings and the scalar variable representing other information.
Since this method incorporates information dynamics, it might be an improvement
over the present value of expected dividends (PVED) model as argued in Penman and
Sougiannis (1996), Frankel and Lee (1998), and Dechow, Hutton, and Solan (1999).1
Furthermore, information dynamics also provide the concrete formulation of future
cash flows then traditional discount cash flows model as discussed in Feltham and
Ohlson (1995).
1
The PVED model uses the present value of expected future dividends to common shareholders based on currently available information as the proxy of a firm’s intrinsic value.
Past insurance literature did not offer a concrete method to evaluate an
insurance company and few researches paid attention to insurance companies’ value
as to discuss their stock prices as well. However, the general supports from the
literature for Ohlson’s method stimulates us to examine the applicability of Ohlson’s
residual income valuation method to insurance companies. We use time-series
models to forecast abnormal earnings and capture the information dynamics. The
resulted value as a proxy to the insurer’s intrinsic value is then used to form P/V ratios
which in turn are used to form a regression model to explain the variations in the
insurance companies’ stock prices. The regression model also includes the
conventional P/B and P/E ratios. The sampled insurers are publicly traded multi- line
companies collected from the 2002 Center of Research in Securities Prices (CRSP)
monthly tapes and COMPUSTAT. Our results will show the applicability of
Ohlson’s model and Lee, Myers, and Swaminathan’s findings in the insurance
industry. This study will also be one of the few studies in examining the intrinsic
value of an insurance company and the variations in insurance firms’ stock prices.
We find the intrinsic value estimated by residual income model with six
months abnormal earnings’ forecasts is a little higher than the book value but a lot
lower than the stock price. Ohlson’s method hence does not produce a good proxy
under various measures for the discount rate. Although our regression model seems
to have high explanatory power for the variations in the stock price, the improvement
from substituting the P/B ratio for the P/V ratio is immaterial. Therefore, the good
results found in Lee, Myers, and Swaminathan (1999) do not appear in the insurance
industry.
The remainder of the paper is organized as follows. Section II discusses the
meaning of financial ratio and Section III introduces the residual income valuation
model. Section IV describes the data and constructs the portfolio for the industry
index. The regression modeling and results are in Section V. Section VI contains
