No. of Recommendations: 12

Hi FC,

This post completes my excursion into Ben Graham's approach be estimating the fair price for the Aggressive Port company stocks through the DCF (discounted cash flow) valuation method.

Links to prior posts are as follows:

BG Contest Preliminaries - http://boards.fool.com/Message.asp?mid=19054507

BG Contest 2nd Round - http://boards.fool.com/Message.asp?mid=19106126

BG Contest Conservative Port Winners - http://boards.fool.com/Message.asp?mid=19226420

BG Contest Aggressive Port Winners & HE Valuation - http://boards.fool.com/Message.asp?mid=19233267

BG Conservative Port valuations for PHM and HVT - http://boards.fool.com/Message.asp?mid=19248201

BG Conserviative Port valuations for RS and KWD - http://boards.fool.com/Message.asp?mid=19297594

The companies that are members of the Aggressive Port are:

Skywest (SKYW)

Standard Pacific Corp (SPF)

Building Materials Holding (BMHC)

Burlington Coat Factory (BCF)

Commerical Metals (CMC)

A variety of DCF models can be used and several factors have to be considered before doing a DCF analysis. This is described in Dr. Damodaran's book on valuation. So, I will go through some of those factors here.

[Note: Here I've already made the choice of using a DCF valuation approach over other valuation methods. My reasoning is that I wanted an absolute estimate to compare with the market prices of stocks selected by Ben Graham's relative criteria (Price-to-earnings and price-to-book). Since, the BG companies are relatively stable with postive earnings and cash flows this is do-able. Further both the Ben Graham method and the DCF method assume that the market can make mistakes, but that these will be corrected over time; and finally with a DCF approach you're assuming that the company will be around for a long time...that's consistent with the Ben Graham approach too.]

Back to the issues to consider in DCF analysis:

1) Can cash flows be estimated? If they can't be estimated then you should use a dividend discount model, otherwise you can use free cash flows. Banks are an example where free cash flows can't be estimated nor can reinvestments be easily identified.

Free cash flows and investment needs can be easily estimated for all companies in the Aggressive Port, so I will use a discounted free cash flow in all cases.

2) Stable Leverage: The free cash flow to equity (FCFE) approach and the free cash flow to the firm (FCFF) should yield the same value for equity if all of your assumptions are consistent. Note that the equity value in the DCF based on the free cash flow to the firm can be found by subtracting the value of all debt from the resulting firm value. However, if the firm's leverage is stable you can simplify estimating the value of equity by directly using the FCFE version of the DCF model. If it's unstable then you're better off with the FCFF alternative and making the adjustments you need for the firm's debt.

While there will always some changes in the debt to equity ratio, by and large the companies in the Aggressive Port have stable debt-to-equity. The exception is that Skywest has recently increased it's debt substantially. However, this is due to the fact that it's currently in the process of updating its fleet of aircraft; and the ratio is still well within the BG criteria. Therefore, I've chosen the FCFE approach.

3) Is the current growth rate less than the overall rate of economic growth? If the growth rate is less than the economic growth rate you should use the stable growth DCF model. It exceeds the economic growth rate than a multi-stage model should be used...you should never use a stable growth model with corporate growth rates that exceed economic growth ...unless you really want to delude yourself. Over long periods of time the average US economic growth rates have about 5-7%.

The BG companies are either in stable growth or ones without significant competitive advantages. So, I've either used a stable growth or a 2 stage growth model.

4) If you're using a multi-stage growth model, then consider if the firm has a competitive advantage or not. If not, then use a 2 stage model with a limited period of high growth followed by a low growth stable period. If the firm has a competitive avantage then you can use a 3 stage model with a period of high growth, then a transition period and finally a stable growth period of low growth. If it's a strong competitive advantage with very high short term growth then you may want to use more stages of high growth.

5) Are the firm's earnings positive and normal? If so, then you can use the current earnings as the initial value for your model.

All of the BG Aggressive Port companies have normal and positive earnings. [Note: that in the Conservative Port Reliance Steel (RS) didn't have normal earnings and a normalised value was used in that estimate.]

6) If the earnings are not normal, then is the cause temporary (such as is the case with a cyclical company)? If the cause is temporary you can use normalised earnings based on an average or normal value for net profit margin.

7) If the cause is not temporary, then is the firm likely to survive? If it's likely to survive then introduce an initial stage in which the company's margins are "nursed" back to normal.

8) If the firm is not likely to survive then a DCF approach is not appropriate. You could use either a liquidation valuation approach (if there's not a lot of debt) or an option to liquidate if there is a lot of debt.

Having gone through that you will arrive at a reasonable choice for a DCF model. Depending on the version you've selected there are further estimates needed as inputs to the models. Since I've narrowed my choices to the FCFE model with either a 2 stage or a stable growth DCF model, the additional assumptions I need are:

a) Returns on Equity (ROE) for a high growth stage and the stable growth stage. [Note: In a transition stage the ROE would decline from the high growth ROE to the stable growth ROE.] If the company's high growth ROE is greater than the industry average you should rather use the industry average unless there's a good reason not to. [An example where I didn't use the industry average was Hawaiian Electric (HE) simply because it has been a low-stable growth company for some time and it's ROE has been fairly consistent over that time.]

b) Growth rates for both the high growth (if applicable) and stable growth stages. [Note: In a transition stage the growth rates would steadily decline from the high growth value to the stable growth value.]

You can either assume a value for growth and determine what reinvestment is required based on the formula:

Required retention rate (i.e. % earnings retained for investment) = growth rate / ROE

Or you can estimate the average retention rate (i.e. reinvestment / earnings) from historical financial data and determine growth from the formula:

Growth = Retention rate / ROE

[Note: that with the FCFE DCF model leverage is stable and that the above formulae are valid in that case. It becomes somewhat more complicated with changing leverage as you have make decisions about how much reinvestment is equity financed and how much is debt financed. In any event you will need to assume that the debt-equity ratio is stable during the stable growth period...duh, imagine that.]

Mostly I assume a growth rate for the stable period based on the overall long term economic growth rate based upon the size of the company and it's long term prospects. For the near term high growth period, I usually work it the other way around by estimating the normal reinvestment rate of the company. Exceptions would be if the company has announced expansion plans (e.g. JOS Banks) that differ from the company's historical reinvestment or if the estimated growth rates are much higher than with the average projected growth rates from analysts (i.e. there's some indication that investment opportunities in the future might be lower than the past)....obviously if you think opportunities will be greater than the analysts suspect...great...if your right.

Naturally earnings should be adjusted to remove exceptional items...well one's that are really and truly exception...and not the routine, exceptional items that some companies favour.

Also, along with Dr. Damodaran I treat R&D as a capital expenditure, so only an amortised value of R&D is treated as an expense. The rest treated as a capital reinvestment. However, this isn't an issue for any of the Aggressive Port companies, so I won't deal with it further here.

Finally, there's that pesky discount rate. I use a modification of the Capital Asset Pricing Model (CAPM) with Dr. Damodaran's bottom's up value for beta.

The CAPM estimate for the discount rate (DR) is:

DR = risk free rate + beta * market premium

The value of beta is based on the correlation between a company's stock price and the overall market price. Mathematically beta is the ratio of the covariance between the stock price and the market price divided by the market's variance. Justifiably it has been much criticised as beta, like any other estimated statistic, has a standard error and this standard error is often so large as to render the estimated value meaningless.

Dr. Damodaran suggests an alternative, the so-called "bottoms-up" beta estimate. Industry betas have much lower standard errors and are, therefore, more reliable estimates. This makes sense as it removes specific company based factors from the estimation. The question is how do you extrapolate from an industry wide beta to an appropriate estimate for a given company. The good Dr. suggests, as a reasonable approximation, that debt leverage is one of the primary factors in determining how well a company does (and consequently it's stock price) in response to the overall economy. This also makes good sense as a highly leveraged company will see it's earnings dwindle as interests rise, but will gain significant benefit if interest rates drop. Of course, there are many other market factors that will influence both a company and its industry, but in many cases it's simpler to use the Dr. D's estimate for beta to come up with a discount rate and than see how sensitive your final price is according to those changes. Given the errors associated with all the other inputs, it's probably not worth more fine tuning. After all, the market premium is known to vary from about 0% to over 10%. How's that for uncertainty.

Anyway, the formula for the "bottoms-up" beta is:

company_beta = industry_beta * (1 + (1 - tax rate) * debt / equity)

For the other inputs to the discount rate, I'll use 5.5% for the risk free rate which is close to the current rate and 4%-5.5% for the market premium. The long term historical value is about 5.5%, but this includes the long hang-over from the Great Depression. The value of 4% is what I estimate is the current market premium based on market fundamentals for the S&P 500.

The covers the 1-2 page primer in DCF models...just enough to confuse you, I'm sure. If you are so inclined though to sort it out, I suggest the Dam Course posts on this board, which has links to worked examples and Dr. Damodaran's own web site.

With that I've gone ahead and estimated DCF estimates for the five Aggressive Port companies. I've summarised the inputs to the DCF below and the results as well in table. I haven't bothered to go into all the issues for each company as my intention is not an analysis of each of these companies per se but to compare two ways of finding undervalued / safe stocks....the Ben Graham method and the DCF approach.

The basic estimated price is based upon a market premium of 5.5%. The range is estimated by using a 4% market premium for the higher value and a stable growth model for the lower value.

The abbreviations used are:

EPS = Earnings per share

FCFE = Free Cash Flow to Equity (i.e. free cash flow after investment for growth)

DR = Discount Rate

GR = Growth Rate

LT = Long Term

RET = Retention rate (of earnings for investment in new projects - i.e. growth)

ROE = Return on Equity (as per the Damodaran course this is return on TANGIBLE equity)

ST = Short Term

The company inputs are as follows:

Company SKYW SPF BMHC BCF CMC

EPS $1.32 $3.71 $1.41 $1.51 $1.43

Model 2-stage 2-stage 2-stage 2-stage stable

ST ROE 35% 18% 12.5% 13.5% 11.5%

ST RET 50% 50% 95% 85% 50%

ST DR 12.5% 10.0% 13.9% 10.7% 11.7%

LT ROE 10% 17% 12.5% 11% 11%

LT GR 4% 4% 4% 4% 4%

LT DR 11% 10% 11% 10% 10%

Price $26.76 $56.66 $14.04 $21.81 $17.20

Range $20-27 $50-60 $14-15 $16-22 $17-18

Buy Price* $13.61 $30.76 $12.46 $17.60 $16.08

Current $18.36 $34.69 $12.81 $18.36 $18.00

* This is the closing price as of 1 May 2003, which is the date the companies were selected via the Ben Graham criteria. The current price is the 30 July closing price.

The gain in the last 3 months has been 12.7% in price appreciation. I haven't kept track of the dividends, but the annual yield based on 1 May prices was 1.1%, which equates to about 0.3% quaterly. That gives a total yield of 13.0%.

In comparison, a US index mutual fund (TD US Index in $US) has had a return of 8.0%. I've chosen an index fund as it includes transaction fees and dividends automatically and makes a suitable comparison for an investor.

For interest, the Conservative Port group has had a return of 12.7% and with an dividend yield of 2.3%, this gives a yield of 13.3% over the last 3 months.

In conclusion, it appears that the Ben Graham method does identify undervalued based on a DCF valuation of selected companies. In this case the selected companies has a group may be 20-30% undervalued - with the lower figure based on an assumption of stable growth.

Of course, each of the companies has a current issue associated with it...in the case of SkyWest it is it's business arrangement as a regional carrier with United, which is now reorganising under Chapter 11. By the way at it's current price SKYW would not meet the Ben Graham criteria; and perhaps should be evaluated to see if it should be replaced in the portfolio. By aggressive, BG meant that the investor should have to spend more time watching over the each company. No doubt this is due to the lower PE threshold, which usually implies that the market does have some concerns over each company. Put another way, considering the issues surrounding SKYW is a reasonable investment with an earnings yield of 10% (based on a 1 May price) and is it still a reasonable investment with a 7% earnings yield based on yesterday's closing price...keep in mind that the risk free rate is about 5.5%.

I found this an informative excursion into BG as I had thought his criteria old-fashioned and not as relevant to today's market. Between BG and me, one of us was wrong. Can you guess which one?

Cheers

SNS