Testing the model
Recently, as regular readers will probably have noticed, I've been taking a subtly changed sort of approach in my blog posts. Most of this comes down to red, whose approach I have shamelessly plagiarised in the pursuit of self-improvement and better stock picking. There hasn't been a monumental shift in the way I discuss things - there never was going to be - but there is a change of direction in how I calculate and appraise investments on my end. In the simplest terms, I've just been shown a much better way of trying to put abstract concepts into numbers. Check out his blog if you want a far more nuanced and less clumsy explanation/application of his way of calculating 'value', but I'll bludgeon a simplistic version of it in four bullets here, with the proviso that any glaring oversights are entirely a result of my own misinterpretations:
- Look at history to come up with a reason cycle-average returns figure for the company in question, making sure to adjust for accounting quirks, exceptionals (?) etc.
- Come up with an operating capital figure. This takes into account leases, your calculation of working capital, and PPE in use.
- You now have a realistic adjusted return on capital figure to compare to a reasonable estimate of cost of capital. The comparison of these two figures is informative - the directionality and size of the gap guides further research.
- Given an assets figure and a return vs. CoC figure, you can calculate the value of the cashflow sans growth. After you adjust for debt/debt-equivalents and tax effects, you have a value figure for the business to compare to the market cap.
Similar to how the DuPont formula guides our analysis of returns - it helps us breakdown the causative factors of a change in returns by splitting a complex figure into easier chunks - by coming up with a value figure with logical and intuitive steps, we can more easily see how things might potentially change. It strikes me as sort of like a DCF, only with the focus on the inputs instead of the outputs.
That introduction is important, because it frames my discussion of Enterprise Inns nicely.
Enterprise Inns is a pub group - the largest leased and tenanted pub group in the UK, with 5,902 pubs. Like many of the other pub groups, they saw their market cap plummet through the recession, as both trading and the bullish valuation of their estates take a nosedive. That's an important point to remember - much of the value in a company like ETI is in its estate and not its operations. In fact, they hold £4.3bn worth of property - far larger than the £528m market cap. It should be obvious from that previous sentence that there's some hefty leverage in play, then. Indeed, the group holds about £2.7bn of term debt and a small amount of bank borrowings. It's a big leveraged play on two things, then - the performance of the operations, the pubs, and the valuation of said pubs. In reality, those two aren't really that distinct - if pubs are performing well, their valuation should hold up. If it's impossible to make money running a pub, the values should tank. That's just a microcosm of exactly the sort of investment framework I'm looking at Enterprise Inns under.
The leveraged and more diverse nature of the returns, then, leaves me struggling for how best to approach the valuation. Do I calculate returns post pub revaluation? This has the effect of inflating pre-recession figures and depressing the last few years. On the other hand, considering revaluation as post-operational noise (which it might well be) materially alters our valuation. It should only be left out if we think it is reasonable to assume the book value of these properties will stay flat going forward. What would be a mundane question is made critical by the size of the differences.
The question is complicated by issues of cyclicality and long-term trend. The ideal scenario would be to take a cycle-average RoC, post-revaluation movement, since that seems the neatest way of looking at their historic returns. Given big secular shifts in people's consumption habits, though - people simply don't use pubs like they used to, and there are generational issues at play - even this is inadequate.
Even leaving that issue aside, we have more interesting altering considerations. ETI's returns have dropped year on year for the last 6/7, and are now at a point below any reasonable estimate of cost of capital. This means two interesting things in their case. Firstly, it means (paradoxically) that contraction is 'profitable'. This is the inverse of the case where a poorly-performing company is expanding; if a company is earning below its cost of capital and chooses to continue pumping assets into growth, that growth reduces the value of cashflows to investors. Investors would've been better served by the company just returning operational cashflow via dividends or buybacks. Since ETI is selling pubs, they should be increasing shareholder value. The practical application of that theory is a little more convoluted, since the paper trail in the annual report is confusing and the accounting treatment seems to obfuscate the information I'd really like to know - what price is achieved for pubs that have been put up for sale, relative to their book value just prior to them being considered for disposal?
There's more points of interest, but since I'm hitting my self-imposed word limit (on a post notably more dry than most, no less) I'll save them for later, if I come back to ETI.
No inkling of valuation since I don't know how to treat several issues with ETI, but hopefully it serves as a reasonable demonstration of how to confuse oneself with figures. It's also a neat way of showing how things I used to take for granted - like PTBV - are more complicated than they might be at first glance. There's all sorts of issues, from the of valuation to the use of those assets and to try and separate book value from operational value is impossible - the two go hand-in-hand.