I've been spending some time going through the Thrifty 30 portfolio curated by Richard Beddard's alter ego, The Human Screen. I thought it might be interesting to post graphical representations of the value-and-price trajectories of each member of the portfolio -- x-rays that illustrate the different ways in which Mr. Market catches up with, and runs away from, reality.
UPDATE: Value = "normal ROIC"/R x Invested Capital, where R = 4% above the after-tax cost of debt.
Forward growth is built in. Equity = EV less after-tax non-operating liabilities plus cash. Leases have been capitalized and pension liabilities taken into account.
Here's the first batch:
UPDATE: Value = "normal ROIC"/R x Invested Capital, where R = 4% above the after-tax cost of debt.
Forward growth is built in. Equity = EV less after-tax non-operating liabilities plus cash. Leases have been capitalized and pension liabilities taken into account.
Here's the first batch:
Crikey, what am I looking at? Green is value, red is price. How do you define value?
ReplyDeleteValue = "normal ROIC"/R x Invested Capital, where R = 4% above the after-tax cost of debt.
DeleteForward growth is built in.
Equity = EV less after-tax non-operating Liabilities plus cash.
Interesting. I use almost exactly the same formula. I have found it goes badly wrong when leases enter the picture though, I am not too sure about subtracting the full value of leases/pensions (always better to be conservative though).
ReplyDeleteI have been thinking about doing a very similiar thing as above, I haven't worked out prices yet though. The complication with fiscal year ends and when results are released has always seemed a bit tricky. What are you doing?
Hey -
ReplyDelete1st question: operating leases are made up of two components: (i) the value of the equipment, building etc; and the interest on the financing. If you add back the entire amount of the operating lease, it will always go wrong, and badly so if the operating leases are a significant share of the total assets in place - as in the retail & airline sectors. The thing to do is to add back the implied interest expense: if EBIT is before interest expense, it should also be before lease interest expense; the remainder is the analogue of depreciation and should not be added back. As a general rule of thum, I capitalize at 7x this year's rent or next year's minimum lease comitment and add back 35% as implied interest.
I recommend the book by Koller et al as an excellent guide to method re: pensions, leases, taxes, securitized receivables etc
On the 2nd Q: I'm interested in prices that are way off, so for past years I generally use the average price in the month of the company's fiscal year end. I add today's price at the very end but don't have a corresponding value data point -- so if last fiscal year end was Dec 31, 2011, I show average price in Dec 2011, and then I add today's price at the end. I don't worry about TTM data: I want to force myself to look at the big picture.
Yes, I have used the same book. My point wasn't on the adjustment it was about the effect on taking it away on valuations. I think that method (although it is the only one I use myself) has an inbuilt bias against leverage, it is just as obvious with normal debt tbh.
ReplyDeleteFor example, if you have a firm making ROIC 10% on $100 of IC, the EV is, with a 10% DR (1x IC multiple), $100. So far so good, this value works in most cases and is a neat way of working out debt capacity.
However, if the same firm was leveraged with $90 debt paying 5% for an ROE of 55% (0.1+(9*0.05)) then the value implied from doing the same calculation on equity (i.e. 10*(0.55/0.1)) is $55. The same value implied by doing the EV-Debt is $10. So one way suggests you should pay 1x book, the other suggests 5.5x book. I don't think one answer is wrong or right but, in practice, the gap between market cap and "value" gets very big with leveraged companies due to this effect...I think. Any thoughts?
About prices, that sounds like a good idea. My fear with using FY end is obviously that results won't have been released until well after that date. One idea I had is to use annual high/low however, this doesn't solve the problem with HY/Quarterly changing expectations in the short term.
Calum -
ReplyDeleteOkay, I think I'm understanding your question. This is how I'd approach your example:
Case #1: EV is $100 and so is equity
Case #2: EV is $181 and equity is $127
because, in Case #2:
ROIC/R = 10%/5.5% = 1.81
IC = $100
Therefore EV = 1.81 x $100 = $181
less debt of $90 and adding back tax shield of $36 at (40$ tax rate x $90 debt) = $127 of equity.
Turning to book multiples: book is $10 in Case #2, not $100. So the book multiple is 12.7x.
In practice, of course, 90% leverage at 5% cost of debt is not possible.
The more usual case is 50% at 5% cost of debt for a business with predictable cash flows. In that case,
Cost of capital = weighted average of (1/2 x 5%) and (1/2 X 10%) = 6%
EV multiple = ROIC/R = 10%/6% = 1.67
1.67 x IC = $167
less $50 debt
plus $20 tax shield
= $137 equity
Book is $50 so the book multiple is 2.74x
That's my take on it and I think its consistent with the view that benefits of leverage can only come from reduced cost of capital and from the government's contribution in the form of the tax shield.
--
I think you're right that results release month is better than FYE month.
Took me a while but I think I see what your saying. However, I don't quite see why the outcomes are different. In particular, the implied earnings yield changes with the cost of debt which it doesn't do in the Equity/ROE model (although the book multiple does change. For the same amount of earnings and the same ROE it suggests paying a higher price (assuming no taxes). Interesting stuff though.
ReplyDeleteBtw, I put one together with yearly high/low ranges, gives a new perspective on valuing growth I think.
Well, in EV approach (using your 90% debt case):
DeleteCash flow from the business is $10
Interest payments to the bank are $2.5
The government gives you $0.2
Therefore equity earns $10 - 2.5 + .2 = $7.7
We used 12x as the book multiple so the yield = 7.7/127 = 6%
And 6% is the cost of capital, so that makes sense.
This seems very similar to an APV approach, where EV is calculated and then debt + tax shield effects are added on at the end.
ReplyDeleteIs "R" an approximate for equity cost of capital? And what's the logic behind cost of debt + 4%? 4% roughly the equity risk premium?
And couldn't this be substituted for a hurdle rate, like 10%?
Hi -
ReplyDeleteThe only difference between this approach and APV is that APV uses cost of equity rather than cost of capital. The value of the APV approach is highest when capital structure shifts around or is expected to shift around; the tradeoff is that cost of equity is likely to be a little too conservative in many instances.
I use R to mean the discount rate which, in this post, is the weighted average cost of capital. If there's no debt, then R = cost of equity.
I use r + 4% as the equity risk premium in situations where I believe there is relatively little fundamental or competitive risk to the business model of the company that I'm thinking about.
I am usually not so interested in valuing or investing in businesses that have no sustainable ability to generate economic profits -- small differences in estimated R and estimated ROIC could see one proved dead wrong about the value of such so-called "commodity businesses".
But, if a company has predictable cash flows and its long-term creditors are comfortable lending to it at, say, 5%, then I think it likely that the implied cost of equity be 9%%.
Because I generally don't go for "commodity businesses" or "contrarian plays", or "story stocks", that 4% risk premium has proven to be a handy, pretty accurate guide over the medium-to-long term.
Finally, I don't necessarily use the cost of capital -- or the cost of equity, for that matter -- as my own personal hurdle rate. My own personal hurdle rate is the expected (annual) return on my most flimsy stock holding: I usually have 4 or 5 positions and 10%-20% cash; if I expect my flimsiest position to return, say, 15%, then that's my hurdle rate.
Very good discussion and appreciate the analysis/response.
DeleteDo you then translate the equity value into an "expected (annual) return"? If so, how?
Or are you just looking for a margin of safety, and figure if the stock can trade up to your expected value in the next x periods, then you end up with a satisfactory annual return?
Btw, hopefully this info might be useful in a future blog post :)
Expected Value
DeleteI believe in market efficiency over the length of a business cycle - 7 to 10 years.
So, if a stock is priced at 100 today when its true value is 200, I think one can reasonably expect that the stock price will appreciate by an average of 10.41% a year until that gap between price and value is narrowed to nil.
That's if value remains static. If one invested in a business that is growing profitably, however, value itself will appreciate and add some % points to that stock price appreciation.
So, for me, the best possible situation is to find oneself invested in a stock that has four features, in rank order:
1. ROIC > R
2. Return on Incremental investment (RONIC) > ROIC
3. g > inflation
4. Price < Value
If these four conditions exist, then time is on your side. That's what Munger taught Buffet and what Buffett means by "a wonderful business".
Margin of Safety
I take margin of safety to mean protection if I'm wrong, not, as I often see it used, to mean "30% or 50% of my estimate of value".
If I buy a stock because I believe that the underlying business is a franchise (item 1 in the list above) then that's the thing I could be wrong about. If the buy price is at net asset value, however, I won't lose money, even if I'm wrong.
If I buy a franchise stock because I believe that its got significant profitable growth ahead of it (items 2 & 3), then a safe buy price is the no-growth franchise value because I could well be wrong about growth.
If I but a stock because I think that it is a viable business -- "has going concern value" -- that's what I could be wrong about and so a safety price is liquidation value.
Sometimes, and in some places, a stock sells at NAV or close to it even though it is an easy to understand, profitable, growing franchise. Coach, in March 2009, for example. In that kind of situation, one's margin of safety is 200% above the prevailing stock price.
--
You're right -- I should type this up in some form of organized fashion and post it. Thanks for the questions and comments.
Red,
ReplyDeleteI am still making my way through your blog (thoroughly enjoying myself along the way) so please forgive me if my questions are answered elsewhere.
I've been trying to research/understand your method of valuation and was wondering if it falls under the EVA methdology. I am a complete novice and may be speaking above my head, but are when you discuss and use ROIC vs. Cost of capital are you finding the value of an EVA perpetuity and then backing into common equity value?
Do you have any suggested primers (yours or others) on this valuation methodology? Also, I was wondering if you ever use other methods like comps/dcf/epv (maybe epv is quite similar?)
Thanks for your fantastic blog and allowing us readers to learn from how you think!!
Hi
ReplyDeleteThe ROIC approach is identical to the EVA approach in that the content is the same and only the presentation differs.
And, of course, they're all basically trying to answer the question: "how much excess cash will this business generate and what is that stream of excess cash worth today?"
In other words, it's all DCF all the way down.
If you take two companies in the same industry, growing at the same rate, the only difference being that one of them has a significantly higher ROIC than the other -- Costco vs Walmart, for example,or Intercontinental Hotels v Millenium Copthorne -- the higher ROIC business will always enjoy higher multiples.
And that's because it generates more excess cash flow than its counterpart because it needs to reinvest less cash in order to growth the business -- the investment rate is (1/ROIC) x g
-- and can therefore pay out more in dividends (or buybacks).
But I can't emphasize enough the GIGO risk inherent in data entry approaches to valuing equities and investing in them.
The more important issue will always be whether the inputs make sense and whether they are reliable guide to the future.
And answering that will always center on whether the business itself is a plausible one, whether you understand why people buy its products/services, why competition isn't a threat, etc.
In most cases these questions are unanswerable and there's no shame in not wasting your time in trying to do so.
But there will be that small minority of businesses that are understandable and whose future profitability can be relied on. Those are worth valuing.
And those stocks that can be valued don't have to be huge or household names.
Take, for example, NICE Information Service: not a household name but it controls the personal & business credit history market in Korea.
That's a business that's unlikely to go bust or to come under severe competitive pressure within the next 10 or 20 years. You can value that. And high ROIC,long term growth with pricing power etc, so it is probably worth valuing.
Anyone who tries to value a teen fashion retailer, however, is kidding themselves.
The hard part is the stuff in between - i.e. neither NICE Info Sys nor Aeropostale.
The more you hug the NICE end of the spectrum the more likely your models will make some sense.
The problem is that if you screen for low multiple stocks you start telling yourself fibs as to why this or that business is above average and why the numbers in the future will be very like the numbers in the rear view mirror.
In any case, for the valuation arithmetic, I think the Copeland. Koller et al "Valuation" book is an excellent primer.
Having read that, I'd go to the Coca Cola investor relations site, grab their annual reports pre-1986 and try to value the shares in 1986. It's an eye-opening exercise.