Intrinsic Value - what's the best way to evaluate a company

BRAGA41's picture
Rank: Chimp | banana points 15

This is my first post here in WSO and I would like to share with you a work I did last semester for an investment course, in which I had to create a portfolio of stocks. After searching for several methods I used a DCF and based my decisions in some Buffett's theory.
I chose 6 companies that from 15th of April are beating Dow Jones by 8%. The purpose of this post is to get some feedback from you guys!

"Valuation methodology:

Financial stability:

It is the stability of a company's business that determines our ability to predict its future earnings. If a company's revenues and costs vary widely, it is not possible to accurately predict growth. We have to take into account that if a company has a stable 10% increase in revenues but loses 20% in one year, in the next year it will need a 50% growth approximately to maintain the same revenue it would have in case the loss did not happen. So we cannot trust in growth averages as a precise value but as instead an estimation and only for stable companies. The gross margin plays here an important role too since if it is too narrow, when the economy contracts, costs rise, or revenues decrease for any reason, this company will shrink it even more having even in some cases negative gross margins. Another figure we must take into account is the level of debt. Highly leveraged firms are too dependent on the debt markets and if a debt crunch happens, which also deteriorates the economy substantially; it would not be possible for them to refinance themselves in those hard times.

Intrinsic Value (IV):

By knowing the financial parameters that a stable company has, it is possible to do a sort in NASDAQ website for example. After the sort we just have to choose the ones which present the more stable and higher returns. By developing an excel sheet that extracts data from Reuters and calculates almost automatically (it needs a few manual adjustments), the intrinsic value of each company it is easy to find which companies are undervalued at the current market price.
For the calculations of the intrinsic value, the operating income, depreciation and amortization, capital expenditures, number of shares, current price, total debt (D) and cash and short term investments (C) are needed (these values are the ones we download automatically for our excel sheet from Reuters).
The owner's earnings (OE) are found by adding the depreciation and subtracting the capital expenditures to the operating income. These earnings represent the amount of money the owner of the business has in "his pockets" at the end of each year. The growth rate (g) from these earnings must be stable so that it is possible to make a prediction of future growth rates. After applying the growth rate to the owner's earnings after several years, there is the need to discount this result to the Present Value. For the discount rate (r) we use 10%, Even though current interest rates are so low that we cannot invest our money at this rate in a risk free instrument (in the best scenario we would get 4%) we use 10% because it gives us an extra safety margin.
Our temporal choice for our investments are 10 years so in these 10 years the stock value will get the returns created by the business, so by summing the PV of these returns we have our first part of the intrinsic value.
The capitalization rate is the returns divided by the market capitalization of the company (M). This way it is possible to calculate given the future returns, what will be the market price 11 years from now. This will be the residual value of the business which is then discounted until the present.
The intrinsic value will be then the sum of the PV of the 10 year returns and the PV of the residual value in the 11th year with the Cash and Short-term investments minus the Total Debt. As recommended by Benjamin Graham, a 25% margin must be discounted from this intrinsic value.

If we wanted to buy the whole business we would need to be worried about the shares dilution, however as we intend to have small stakes of each of these companies in our portfolio we need to take in account this variable. By analyzing the number of shares in the past 10 years of the company's activity it is possible to make a prediction, with the respective safety margin, of the future dilution rate. So when we calculate the intrinsic value per share we can already take into account this factor and thus arrive at a more precise value.
The conclusion is simple, if the current price of a stock is lower than its intrinsic value per share then we consider it as undervalued.

Fundamentals:
It is not enough for a stock to be undervalued. It is imperative that the firm has some kind of competitive advantage that its competitors do not have and cannot obtain at least on short term. This competitive advantage can be a way of reducing costs by for example having a better bargaining power with the suppliers or increasing sales by producing goods with an extraordinary quality. A firm which belongs to a competitive industry must have some factor that differentiates from the others because otherwise it would be affected by price competition and narrow margins. The management quality is another aspect to take into account, which can be evaluated by the precision of their estimates and the transparent financial results and accounting practices. Customers and suppliers are also extremely important since a company that relies in a small number of suppliers and customers is dependent on their stability."

The companies inside my portfolio are: SNX, CHRW, JW-A, WMT, KSS, MNST.

Please feel free to give me some advice regarding the calculations and methods used to find the Intrinsic Value.

AttachmentSize
PDF icon Stocks report.pdf653.79 KB

Comments (25)

Aug 21, 2013

I didn't read your post, but if you want to find the most accurate estimate of intrinsic value, just call up the company's CFO and be like, "Brah, I know you possess third party estimates of your company's intrinsic value, you gonna give those to me?"* After he says no, don't invest, that company is too hard to work with.

*This is a proprietary approach, please don't share with anyone else.

    • 1
Aug 21, 2013

thats not that simple of a question. one way, however, is to look at what the company would make if it sold everything and payed out all other obligations.

Aug 21, 2013

Why wouldnt you just do a comp analysis of its industry peers? There are only so many ways to value a company, all of which provide quick and dirty estimates.

Public comps, transactions comps or DCF

Aug 21, 2013

liquidation value is a pretty poor proxy for the ongoing value of a company as it is likely a lot lower than its true value as its assets would be sold at fire prices.

comps are the best way. look at multiples (EV/09 consensus EBITDA, P/E, etc). if its peer group trades at 7x and the company trades at 6.0x, you need to determine if its undervalued or if there is a reason it should trade at a discount (slower growth, lower profitability, etc)

Learn More

Side-by-side comparison of top modeling training courses + exclusive discount through WSO here.

Aug 21, 2013

In the example of Caribou Coffee, a company I am researching, they actually don't list he P/E ratio

Aug 21, 2013
MandA_Junkie:

In the example of Caribou Coffee, a company I am researching, they actually don't list he P/E ratio

UHH That is because their 2007 actual earnings is negative, and analysts' estimates for future earnings is negative as well.

In this case, you should use EV/EBITDA.

And take a closer look at your company. And your Finance textbooks as well.

Aug 21, 2013

ev/ebitda is better in my opinion as earnings can be full of a lot of "junk" that isnt true earnings ie non cash items. if you can get onto a bloomberg, you can get consensus 2009 EBITDA estimates, if not im sure they are out there somewhere.

and sorry to insult you if you know this but EV= market cap (# shares outstanding * share price) + net debt (total debt-cash)

hope that helps

Aug 21, 2013

comps can strongly reflect market sentiment, so in a lot of cases it can't provide much insight

of DCF models i've looked at, i like CFROI. it seems to most reasonably model a lot of key variables better than some of the other models that look mathematically elegant but explain reality poorly (ie, most quick and dirty models).

also, accounting effects deserve strong consideration

Aug 21, 2013
Ivar Kreuger:

comps can strongly reflect market sentiment, so in a lot of cases it can't provide much insight

I think valuation comps can provide plenty of insight. To account for current market sentiment, you would need to look at comps over time to see how things are trending.

Companies tend to directionally track their peer group over time so a divergence in comps can indicate material changes that might be worth drilling down on.

Aug 21, 2013

Let what you think the shares are worth = x
Let current price = y

Upside = x / y - 1
Discount = 1 - y / x

What you think the shares are worth is a whole other question.

Aug 21, 2013

If fair value estimate is 5 and price is around 2.8 would that make it undervalued by 55%?

Aug 21, 2013

1 - 2.8 / 5 = 44% discount

Learn More

Side-by-side comparison of top modeling training courses + exclusive discount through WSO here.

Aug 21, 2013

bump

Aug 21, 2013

Yes, you take the target multiple (here 15x) and multiple it by the homework stock's EPS. With a range of valuations you would just multiple the eps by the range of multiples and, taking other factors into consideration, try to decide where in that range homework stock should fall.

In practice it's rarely that easy and using current market trading multiples doesn't always yield "intrinsic value". First, there are usually good reasons for different multiples such as different margins or different growth profiles. These factors are very important. So, if homework stock has half the margins and is shrinking while the 15x industry is the growing, it would be a poor decision to apply that 15x multiple thinking that it would / should trade at that level. For example GOOG multiple is about 1/2 of BIDU. While they're very similar, BIDU has higher margins, much stronger growth, and major exposure to china (which could be good or bad depending on your outlook). So one might say that their respective multiples are justified.

In terms of intrinsic value, I find that using buyout / take private multiples is a more effective way to determine intrinsic value.

    • 1
Aug 21, 2013
  1. The DCF model takes into account a terminal growth rate (g). Terminal value of a company = (FCF at year 10) *(1+g)/(WACC-g). In other words, the DCF model spits out the present value of a company's cash flow from now to the rest of eternity, assuming a growth rate of g (usually assumed to be 2.5%, aka long term economic growth). Also, the DCF assumptions are never set in stone. You are supposed to regularly update the assumptions based on new information.
  2. DCF is faulty when you're trying to value companies with inconsistent/difficult to predict CF. This is when you need to use other valuation metrics, like multiples. One such example is valuation of banks/financial institutions.
Aug 21, 2013

You need to keep in mind that NPV you get from just the cash flows needs to be adjusted for net debt. In your example, the NPV of the cash flows doesn't change, but the company has more cash after year 1, so the value goes up. DCF will only give you the value of the company as an operating entity, and won't include things like cash on the books, debt on the books, etc.

You calculate the NPV for the cash flows but then adjust it for the value of FINANCIAL non-revenue generating assets (i.e. cash and debt). You don't use current book value, because some of those assets are revenue generating assets and you would be double counting (i.e. counting the value of a factory's cash flows, and counting the value of that factory in the assets; this is double counting because the inherent value of the factory is tied to its ability to generate a cash flow stream).

There's lots of little nuances to this shit, ranging from whether you do cash flows to equity or cash flows to equity and debt, what adjustments you make as far as getting to a free cash flow, what discount rate you use and how you arrive at it, etc. It's art just as much as it is science (something that's lost on a lot of people in finance).

Aug 21, 2013
  1. Because the present value of the expected cash flows is is the "extra" value and is constant. Say you have a 1 stock company with 0 assets and 0 retained earnings that is expected to make 100 every year forever with a 10% discount rate p.a. . You dont expect the company to be worth 100 the first year and 200 after the second year, you instead see what is the discounted present value of future expected cash flows. It makes sense if you ask yourself how much 'would you ask for such a stock, 100 in year 1? doesnt make sense for there are future expected cashflows.
  2. RIM

Valor is of no service, chance rules all, and the bravest often fall by the hands of cowards. - Tacitus

Dr. Nick Riviera: Hey, don't worry. You don't have to make up stories here. Save that for court!

Aug 21, 2013

Shit. Are there any good books/websites that explain this stuff?

Aug 21, 2013

Change your name

Aug 21, 2013

I am answering because of OPs superior name. Please don't change it.

  1. The stock's shareholders' equity will as you say increase year on year. If you use a valuation model where shareholder's equity is a big input (let's say a Price/Book, a multiple-based valuation), then yeah, the price should increase year on year.

But let's say you think Price/Book is a retarded way to value stocks. For example you might go: "Hey, you know what? I have a money tree with $100 on it today that grows at 10%. Why should I sell that at the same price as another money tree with $100 on it today and grows at 1%? I want to include those different growth rates in the valuation, but how can I?"

Clearly, if you think this way, then you need another valuation model that is NOT Price/Book.

Let's instead try with a discounted cash flow model. This rather takes the cash flows experienced at each point in the future and calculates a present value today. The funny thing about the DCF model is that it doesn't give a flying fuck about what shareholder's equity is at any point. Shareholder's equity could be a gazillion, but unless there's a cash flow attached then nobody cares. After all, if you have something incredibly valuable that never gives you any earnings, then why would anyone pay something for it? If nobody would pay something for it, how is it valuable?

The DCF model specifically would say, for example, income of $10 per year, discounted by (r - g) (i.e. discount rate minus growth rate) for infinity. If r is 5% and g 2.5% then the fair value of the latter money tree is ~$400. Nobody cares how much money is on the tree at the moment. Totally irrelevant.

What about reinvestment? Well, the DCF model will give you a price that is correct TODAY taking the whole of eternity into account. But it's true that if you run a DCF TOMORROW, then the price is higher. Remember the time value of money though - this has already been factored into what is a fair price TODAY.

Lastly, you might feel it's retarded to ignore current shareholders' equity when you value stocks, as you astutely point out. In that case you could choose a different model: Book value plus present value of those future earnings that exceed the cost of capital.

In short, it's pretty much up to yourself how you choose to value anything. People use all kinds of valuation models depending on what they had for breakfast, what everyone else uses etc. It's not like they give the same answers (why would they when some models care about growth and others don't?).

  1. Under DCF it's worth nothing because we presume the only thing people care about is the cash flows you get from something. It stands to reason because why would anyone pay anything for the company if they never make any money from it?

So there's a couple of things you can do. You can use a valuation model that doesn't care about cash flows or growth. OR, you can use a DCF but instead of presuming having no cash flows to infinity (how much would YOU pay for that opportunity?), you put up a cash flow in 1 year's time of "selling everything on the black market at book value" and discount that back to today. Presto, the value is not 0.

Aug 21, 2013
Comment
Aug 21, 2013