WACC - weighted average cost of capital or discount rate

[TheMotorcycleBoy]

My current hobby-horse, as has been noted, is trying to value company shares by reviewing company financial statements and applying "fundamental analysis" techniques. A concept which frequently crops up is the discount rate.

The most recent valuation methodology which I have been researching is "Earnings Power Value", and I'm still really trying to get my head around all it's facets, i.e. not just wack out the formula, prior to writing a thread on this. In my researches so far, I have read up a couple of online documents, for example:

https://seekingalpha.com/instablog/2929 ... estop-corp
https://moneyweek.com/361556/earnings-p ... -has-legs/
(no paywalls obvious in either above)

One of the steps in the EPV calculation process is dividing an "adjusted earnings" figure by a value in order to give an equity value for the company. This value is described in the first document as being the WACC (weighted average cost of capital) and in the second document as the discount rate. After noting the substitution of terms above, I was reminded of a wiki I read earlier, where again, an equivalence between these terms is suggested i.e.:

 b) This equation is also used to estimate the cost of capital by solving for r
    r = (D1/P0) + g

So it seems that discount rate and WACC are indeed two sides of the same coin. One is from the investor's side, i.e. how risky a venture he/she considers the company, and thus "how much reward" is expected for investing in the company, and from the company's perspective - how enticing their interest rate must be, in order for someone to "take a risk" and lend it some money.

To date I have pondered upon how exactly to set an appropriate rate in a model, i.e. my desired rate of return. This is usually described as being:

rate = risk free rate + risk premium

Where an easy to understand value for the risk free rate is the current 10 year gilt yield (no paywall obvious), whereas the risk premium is really down to each investor's appetite etc. and thus the rate can be seen as being a little subjective.

Anyway, seeing as my readings lead me to believe in an acceptance of a more formal equivalence between the discount rate and WACC, I decided to research definitions of WACC, and in a subsequent post, will play around with some online formulae for this value; probably citing my current favourite of Marshalls Plc (MSLH) as an example.

Matt

[dealtn]

Be sure to remember that many companies have different forms of capital. So they might have debt, pref shares, equity shares etc. each of which will have different "cost of capital". The weighted average cost of capital reflects this. This isn't the same as the equity discount rate used to value the equity value of a company.

Take an extreme case where 99.9% of a company's capital is in debt form at a very low interest rate, and the 0.1% is a single share in the company. The WACC is very close to the debt interest rate. The company may be very marginally profitable, with most earnings used to service the debt. It would only take a small drop in profitability for the company to be loss making. The equity is very risky. You would need a high discount rate in any model to reflect this.

[TheMotorcycleBoy]

Be sure to remember that many companies have different forms of capital. So they might have debt, pref shares, equity shares etc. each of which will have different "cost of capital". The weighted average cost of capital reflects this. This isn't the same as the equity discount rate used to value the equity value of a company.

Take an extreme case where 99.9% of a company's capital is in debt form at a very low interest rate, and the 0.1% is a single share in the company. The WACC is very close to the debt interest rate. The company may be very marginally profitable, with most earnings used to service the debt. It would only take a small drop in profitability for the company to be loss making. The equity is very risky. You would need a high discount rate in any model to reflect this.

Thanks - you make a good point. I have got a few scribbles which I was going to post later on today or tomorrow, regarding following up on some more descriptions of WACC.

But regards what you've added here The company may be very marginally profitable, with most earnings used to service the debt. It would only take a small drop in profitability for the company to be loss making this sounds like it's leading to something else I have recently read, and that being ensuring that a firm's ROCE > WACC, if I recall correctly.

thanks again
Matt

[TheMotorcycleBoy]

Be sure to remember that many companies have different forms of capital. So they might have debt, pref shares, equity shares etc. each of which will have different "cost of capital". The weighted average cost of capital reflects this. This isn't the same as the equity discount rate used to value the equity value of a company.

Actually not wanting to get into a big row about it but in these articles the authors actually do use WACC as the discount rate when devising valuation models.

https://seekingalpha.com/instablog/2929 ... estop-corp
https://finbox.io/blog/weighted-average ... acc-model/

I guess the usage of the term equity gets a bit blurry at times?

I suppose as a share-holder (bond-holder or the bank manager) of firm, having a view on the overall cost of the capital (i.e. shares + bonds + loans) is the required for an assessment of holding any part of it?

Anyway I'm pretty new in all this, but an appreciation of WACC seems a prerequisite in understanding EPV calculations, so I will soldier on.

[TheMotorcycleBoy]

Have looked up a few online definitions/formulae for WACC. It is expressed as being:

WACC = Weight of debt in capital structure * Cost of debt + Weight of equity * Cost of equity

That's great but what is the most practical way of estimating this, primarily from the firm's financial statements? The weighted components are reasonably straightforward, they are the relative proportions of debt and equity in the capital structure.

Where the weighted debt in the capital structure:

Wd = (Market value of debt)/(Market value of debt + Market value of equity)

And similarly for equity. In order to reduce my work load, I am going to make some simplifications. For market value of equity I will use the present day Market capitalisation value and for debt I will use the Book values of short and long debt. The simple example Marshalls Plc. MSLH I will use, has no preference stock so only ordinary share capital need be considered.

The cost of debt and that of equity were difficult to figure out. The cost of debt to use in the above formula is

Pre-tax of debt * (1 - effective tax rate)

I plan on deriving this from the company's latest report.

Cost of equity, looks a bit more fuzzy. I've read on this of late, where reference to stock indexes and "beta factors" are made e.g here. I'm not going there initially, instead I'm planning to go it alone, basically I'm planning on using my own total returns estimate for MSLH. Something like this, review the 20 year period Jan 1998 - Jan 2018, and get the average capital growth of stock price, and review the last 5 years of dividends in order to get their dividend yield, then I shall add the two values to get MSLHs rate of return for an investor. Perhaps that sounds bonkers? But it's the most obvious/practical way that I can currently devise.

If anyone is interested here are some of the links I found for purposes of thought provocation:

https://accountingexplained.com/capital ... f-capital/
https://www.thebalancesmb.com/calculate ... tal-393130
https://www.investopedia.com/walkthroug ... /wacc.aspx

To be continued....

[dealtn]

OK. I needed to read the original post more closely.

The EPV method is a cautious method that values the assets (I would say net assets, so allowing for liabilities) and adds on to this the earnings power that those (net) assets might be expected to earn, discounted at a suitable rate. No additional value is attached to any growth in earnings (so no time value of money discounting occurs - which is where it is very important to use the appropriate discount rate).

Because the debt is removed when measuring (net) assets, and the interest paid/received is also adjusted out of the earnings (and because "mostly" you won't encounter companies with as extreme a capital structure in the example I gave) using WACC is an OK approximation I would say in such a model I guess.

I would guess this model would be used when looking for "value" situations and unlikely to throw up many candidates where the equity is priced by the market at below these two components.

Two refinements I would look at should you find such a candidate.

Firstly you have ignored the growth, which is fine and conservative, provided there isn't likely negative growth. Consider as examples newspaper companies with media transferring online, tobacco (arguably) with people giving up smoking (or switching to alternatives provided by other companies), "bricks" shops losing out to "clicks" retailers. Plus no doubt other examples you might think of.

Secondly the asset (and liability) values in company accounts are probably "book value" and might not be reflective of the real value it would "cost" to now create those assets. They might be appropriately adjusted by depreciation/amortisation but you might consider ensuring they are appropriate. Similarly if the assets are property they might be on the accounts at significantly different to market values. Again that is fine if they have increased, and this method is conservative, but that might not be the case.

Hope that is of some help.

My earlier comments about ensuring you use the appropriate discount rate are more relevant to time value discounting the future expected flows either to the company or the instruments themselves, so sorry to have caused unnecessary confusion.

[TheMotorcycleBoy]

OK. I needed to read the original post more closely.
...
My earlier comments about ensuring you use the appropriate discount rate are more relevant to time value discounting the future expected flows either to the company or the instruments themselves, so sorry to have caused unnecessary confusion.

Don't worry mate! I'm still learning and playing around with ideas. All comments are useful.

Since we (the wife and I) entered the field of investing very recently (March 2018), we are very aware of being *possibly?* near to a market peak, and also know of a few folks who got very badly burnt in the dot com years. So I'm very keen on trying to be more astute re. "buying a good price". Hence a lot of these type of posts coming from me....

I'm struggling currently with getting a realistic value of the cost of equity for my example of MSLH, I'm probably going to settle on assuming that the firms stock price will grow roughly with inflation, and add to this their average dividend yield i.e. 2.5% + 3% = 5.5%.

Anyway, I'm done for the day, will post more later.

[TheMotorcycleBoy]

So I'm floundering a little bit on the concept of the "cost of equity". Conceptually it seems to me to be the implied cost to the company that it must bear in "keeping it's investors happy". There's presumably two components to this, one being an onus the stock price gradually rising, and the market's expectations given other environmental factors on that happening, and the yield (payout/share price) of the dividends.

In mainstream investing theory it looks as they are 2 ideas for calculating this:

1. The dividend capitalisation model i.e.

Dividend yield + growth rate

This looks similar to DDM however I personally think that for it to make any sense to firms, some of which may not pay dividends, for the growth rate to be the capital growth rate.

2. The CAPM i.e.

Risk free rate + Beta(Expected rate of return from firm - Risk free rate)

I dislike the above model, since firstly I try to avoid magic numbers in formulae, and it just introduces too many unknowns in my opinion.

I plan to ignore CAPM for a while in my investigations, and I'm focussing on something similar option 1. So for Marshalls I studied the following graph. This suggests, at a guess, an average dividend yield of about 2.7%.

Trying to make a guess at the capital growth is much more difficult. Eventually I looked at historical share prices, at July 1988 of £0.54 and July 2018 of £4.07. Using CAGR this gave me 6.9% over the 30 years.

For my continuing calculations I plan using 2 difficult equity costs, one using the above 6.9% as share price growth, and the other assuming recent average inflation figure of 2.5%. Thus this gives me:

high equity cost = 6.9 + 2.7 = 8.6

low equity cost = 2.5 + 2.7 = 5.2

[TheMotorcycleBoy]

Now to the cost of debt. I'm using the data from Marshalls 2017 report.

Firstly I tried just searching in the document for "effective interest", which took me to Note 16. on page 99:

31 December 2017
================
Bank loans	           Variable	1.97%	43,883
Finance lease liabilities  Fixed	10.0%	   259

We can see the amount (43,883) for the bank loans dwarfs the amount FLL amount, hence we may as well just round up the above to 2%. Out of interest, I did some more digging around. In the income statement we have:

Financial expense (1,388)

and the accompanying Note 5. on page 90:

Net interest expense on defined benefit pension scheme      377
Interest expense on bank loans, overdrafts and loan notes 1,005
Finance lease interest expense                                6

If we ignore the interest due the (closed in 2006) DB pension scheme, and review the balance for the short and long loans (35 + 44107), we can calculate the cost of debt to the yearly running of the business as:

1005/(35 + 44107) = 2.3%

So given my earlier investigation at Note 16. I am confident at using this figure of 2.3% as the pre-tax cost of debt for Marshalls.

The figure to use in WACC is post-tax cost of debt, so another search for "effective tax" in the above report gives me 19.1% from page 27. So post tax cost of debt:

2.3(1 - 0.191) = 1.86%

[TheMotorcycleBoy]

For the calculation of the weighting values, I took the odd shortcut. The online references state "Market value of equity", fair enough I just used the market cap. However for the "Market value of debt", I'm afraid I took the book values, i.e. the interest bearing loans/borrowings figures in the balance sheet. Here goes:

Total equity	829000
Total debt	44100
Total capital	873100
	
Wd	0.051
We	0.949

So combining these weights with the earlier costs of equity and debt, I arrived at the following as the WACC for Marshalls for 2017:

high = (8.6 * 0.949) + (1.86 * 0.051) = 8.3%
middle = (6.9 * 0.949) + (1.86 * 0.051) = 6.6%
low = (5.2 * 0.949) + (1.86 * 0.051) = 5.0%

Note that I used high equity cost (using 30 year geometric average for share price growth), low equity cost (using expected inflation rate these days) and an average/middle value.

Hmm.... not sure exactly where the above calculations get me. Other than to assume that any value between 5.0-8.3% is a half way decent value to use in a MSLH share valuation. I guess using 8.3% will lead to more conservative estimates...

Any comments on the relevance of these values? Do they look like the kind of numbers which analysts would derive?

[cshfool]

I think Terry Smith usually assumes about 8% for WACC - hence the focus on "only buying good companies" whose ROCE greatly exceeds its WACC as part of his odd strategy, as you highlighted in bold. This is good stuff Matt - impressed by these posts and the effort put in. Mr Bearbull at the IC has a spreadsheet with Renishaw as an example that uses a similar methodology to come up with an EPV aka his "simple valuation model", along with PE ratio /growth rate as another model.

I'm not sure if there is an easy answer to the question how much is a share worth in terms of intrinsic value, but the question surely is a valid one.

Colin

[TheMotorcycleBoy]

I think Terry Smith usually assumes about 8% for WACC - hence the focus on "only buying good companies" whose ROCE greatly exceeds its WACC as part of his odd strategy, as you highlighted in bold. This is good stuff Matt - impressed by these posts and the effort put in. Mr Bearbull at the IC has a spreadsheet with Renishaw as an example that uses a similar methodology to come up with an EPV aka his "simple valuation model", along with PE ratio /growth rate as another model.

I'm not sure if there is an easy answer to the question how much is a share worth in terms of intrinsic value, but the question surely is a valid one.

Colin

Hi Colin,

Yes, I'm building up to a EPV for Marshalls, and will write a thread for that sometime. I say 'a', though there could turn out to be several, as I try to refine my technique, as I'm finding several variations upon the theme.

This thread seems to give a brief taster into EPV
https://moneyweek.com/361556/earnings-p ... -has-legs/

Whereas this blog gives a lot more background, and generally seems more plausible
https://seekingalpha.com/instablog/2929 ... estop-corp

...and even better, referred to a nice book on the subject which has jumped to the front of my reading queue! I've already dipped in part way into it, into to an EPV valuation of WD-40 (I believe they do one on Intel, in the next chapter), and I'm trying to base my initial workings on this. However, just like you and I noted here previously, regarding Phil Oakley's book, you do still need to keep your eye on the ball to figure out any jumps in the maths/accountancy work, if you know what I mean.

[dealtn]

Just be aware of a potential issue with cyclical, and perhaps more importantly, seasonal businesses using your "method". Not a criticism, but the balance sheet is a snapshot of a single date, whilst the p/l (or cashflow statement) is reflective of the entire accounting period.

It will be the case that the interest cost covers the period, but applying it to a single date to work out the interest rate, will be inaccurate. Depending on the company, and the detail of the accounts its auditors produce, you might get a note of the rate, the average debt liability, or other additional information (such as debts repaid in the period, having an interest cost, but no mention in the balance sheet) to assist you. The "error" is likely to be small in what is a general valuation method anyway, but it may well be significant so best to be aware of its potential existence.

Some seasonal businesses choose their balance sheet and accounting periods for a reason!

[TheMotorcycleBoy]

Just be aware of a potential issue with cyclical, and perhaps more importantly, seasonal businesses using your "method". Not a criticism, but the balance sheet is a snapshot of a single date, whilst the p/l (or cashflow statement) is reflective of the entire accounting period.

It will be the case that the interest cost covers the period, but applying it to a single date to work out the interest rate, will be inaccurate. Depending on the company, and the detail of the accounts its auditors produce, you might get a note of the rate, the average debt liability, or other additional information (such as debts repaid in the period, having an interest cost, but no mention in the balance sheet) to assist you. The "error" is likely to be small in what is a general valuation method anyway, but it may well be significant so best to be aware of its potential existence.

Some seasonal businesses choose their balance sheet and accounting periods for a reason!

I see what you mean. When I finally get round to posting up my MSLH example you will note that I'll average over 4 years (to normalise fluctuations in A&D and capex), but I take your point - that a construction materials supplier may have different monies in and out throughout a year.

[TheMotorcycleBoy]

This is mainly a note to self, for future investigation:

http://www.economicsdiscussion.net/capi ... pital/4415

but obviously for others to comment if desired!

[dspp]

This is mainly a note to self, for future investigation:

http://www.economicsdiscussion.net/capi ... pital/4415

but obviously for others to comment if desired!

Matt,

You are coming at this from the wrong end of the telescope I think. This is not a criticism of you, just an observation. This is because you are learning outwards from your position as an individual investor. As a result you don't have the structure to be slotting the individual pieces into, and so the rest of us (who are undoubtedly learning along with you) are trying to act as guides. That's just my pedagogical observation.

On finance-oriented MBA courses & high-end accountancy courses they go into the subject of optimal capital structure, i.e. debt vs equity and how much of each. The WACC is a piece of that jigsaw. Here are some links, if you want to go online.

https://corporatefinanceinstitute.com/r ... ity-value/
https://www.investopedia.com/ask/answer ... ucture.asp
https://www.investopedia.com/terms/o/op ... ucture.asp
https://www.accaglobal.com/uk/en/studen ... cture.html

Or alternatively one of the many corporate finance textbooks. Here you go, this may get you thinking and help with structure.
http://catalog.mit.edu/search/?P=15.418
https://sloanbid.mit.edu/resources/15.418.pdf

My textbooks are packed away so deeply I'm not even going to think about it !

Regards,
dspp

[TheMotorcycleBoy]

Matt,

You are coming at this from the wrong end of the telescope I think. This is not a criticism of you, just an observation. This is because you are learning outwards from your position as an individual investor. As a result you don't have the structure to be slotting the individual pieces into, and so the rest of us (who are undoubtedly learning along with you) are trying to act as guides. That's just my pedagogical observation.

Hey....

Don't worry about the observation/criticism etc. I know that any advice you give is well-intentioned, Dave. To be honest, it has (as I'm sure that you can well imagine) been a very steep learning curve for Mel and I in the world of investing these last 8 months or so. I think that we've made a few mistakes, we've fluked a few good things, and we've learned one or two lessons the hard way.

What we are, belatedly, trying to teach ourselves the concept of value....and as you all have probably noticed which of our earlier purchases were too highly priced, and what lower "better valued" prices we should "top up" at would look like.

On finance-oriented MBA courses & high-end accountancy courses they go into the subject of optimal capital structure, i.e. debt vs equity and how much of each. The WACC is a piece of that jigsaw. Here are some links, if you want to go online.

https://corporatefinanceinstitute.com/r ... ity-value/
https://www.investopedia.com/ask/answer ... ucture.asp
https://www.investopedia.com/terms/o/op ... ucture.asp
https://www.accaglobal.com/uk/en/studen ... cture.html

Or alternatively one of the many corporate finance textbooks. Here you go, this may get you thinking and help with structure.
http://catalog.mit.edu/search/?P=15.418
https://sloanbid.mit.edu/resources/15.418.pdf

And thanks again, for this Dave, will chew through some of it and comment some more over the next few days.

Matt

[TheMotorcycleBoy]

Here are some links, if you want to go online.

So looking at this one first:

https://www.accaglobal.com/uk/en/student/exam-support-resources/fundamentals-exams-study-resources/f9/technical-articles/optimum-capital-structure.html

and quoting:

The weightings are in proportion to the market values of equity and debt; therefore, as the proportions of equity and debt vary, so will the WACC.

But when accessing the actual weights, does an analyst always use market values?

My bone of contention is when using market values, firstly for the equity when using the market capitalisation (i.e. market value of equity) we note that for many companies wild fluctuations in markcap in a small time frame. For instance in the case of Focusrite TUNE the share price has gone from 504p (1st sept) to 364p (mid oct) back to 415p (now), and as it has zero debt, is has only an equity component. If we have a firm with some debt and a similar marketcap fluctuation, how can market value of equity make any realistic contribution to a WACC calculation since, given that profitability rarely changes so quickly, why would the perceived risk of lending it money change in comparable step with it's short term markcap?

Secondly if in the case of debt the majority is held by a bank, who don't disclose the market value of a loan, surely to the analyst only the book value is the realistic weight assessment on hand?

As the WACC is a simple average between the cost of equity and the cost of debt, one’s instinctive response is to ask which of the two components is the cheaper, and then to have more of the cheap one and less of expensive one, to reduce the average of the two. Well, the answer is that cost of debt is cheaper than cost of equity. As debt is less risky than equity, the required return needed to compensate the debt investors is less than the required return needed to compensate the equity investors. Debt is less risky than equity, as the payment of interest is often a fixed amount and compulsory in nature, and it is paid in priority to the payment of dividends, which are in fact discretionary in nature. Another reason why debt is less risky than equity is in the event of a liquidation, debt holders would receive their capital repayment before shareholders as they are higher in the creditor hierarchy (the order in which creditors get repaid), as shareholders are paid out last.

In the enboldened qualification of risk. Who is the debt least risky for? Presumably what's meant is least risky for the provider of the capital (the bank, or bond holder), rather than the acquirer of the capital (the company).

What about from the company's perspective? Is equity less risky because the firm don't have to pay divs, or more risky because if the stock is sold, the firm is an easily target for a takeover?

Remember that Keg is a function of beta equity which includes both business and financial risk, so as financial risk increases, beta equity increases, Keg increases and WACC increases.

This term "Keg" has been introduced, what is keg?

Can someone help me further? That is some of my above wafflings, esp. "keg". The article of Dave's that I've linked looks interesting, but it does raise questions. I've got as far as Modigliani and Miller’s no-tax model but got to stop now......day job calls

[TheMotorcycleBoy]

And finally by a process based of creative use of google and one's imagination:

https://www.acowtancy.com/textbook/acca ... d-ke/notes
https://www.accaglobal.com/crsh/en/stud ... apm-2.html

"keg" (Geared) cost of equity?

[TheMotorcycleBoy]

Just chewed over a bit of this:
http://www.economicsdiscussion.net/capi ... pital/4415

(With a particular interest in "cost of equity")

I quite this like one:

Earnings Price Ratio Approach:
The earnings price ratio approach suggests that cost of equity capital depends upon amount of fixed earnings
of an organization. According to the earnings price ratio approach, an investor expects that a certain amount
of profit must be generated by an organization. Investors do not always expect that the organization distribute
dividend on a regular basis. Sometimes, they prefer that the organization invests the amount of dividend
in further projects to earn profit. In this way, the organization’s profit would increase, which in turn increases
the value of its shares in the market.The formula to calculate cost of capital through the earnings price
ratio approach is as follows:

KE = E/MP
Where, E = Earnings per share MP = Market price

And I think that their criticisms, are probably unjustified.....provided sensible averaging is applied.

And this critique of it:

c. Ignores the fact that all the earnings of an organization are not distributed in the form of dividend. However, some part of earnings may be kept in form of retained earnings.

seems bizarre, as all earnings at the EPS level, are diced and sliced in either divs, capex, or retained earnings are they not?