On stage with Claer Barrett at last year’s FT Weekend Festival we asked the audience if they invested in shares directly. To our surprise, two-thirds of the room raised a hand. For starters, who knew rooms had hands?
And the response flew in the face of efficient markets. Stockpicking is impossible. If 85 per cent of US equity funds have underperformed the index over the past decade, what chance has Penelope of Tunbridge Wells?
I have written recently that perhaps there is something deep in the human psyche that believes in active management. Besides, you don’t have to beat an index. Make enough money and who cares?
Plus trading shares is fun. Learning about companies, placing your bets, watching prices move up and down. Everyone, it seems, has a view on the future of Amazon or British Airways or TikTok.
People even buy stocks on a hunch. But no matter how financially challenged, I still recommend doing some basic valuation work to make sure you are not being a patsy.
Professional investors like to keep their methods a secret — and in most cases, thank goodness, they do. Meanwhile, the simplest approaches to valuing a company can quickly confuse. And who has the time, anyway?
Luckily, time doesn’t help much. When I was a young analyst my Qantas model came to 150 separate Excel tabs. I forecasted every seat, food tray, flight attendant and landing slot for the next 20 years. All that to discover only the fuel price and US dollar mattered.
What is more, almost everything I was taught hasn’t really worked if you wanted to make serious money in recent years. Like most brokers, I calculated Tesla to be worth zero. And I would have sold the other magnificent six long ago.
But many valuation rules of thumb still apply. I have always paid up to 15 times earnings per share for a top-class bank, no more than eight to 10 times for a crap one. JPMorgan’s share price of $281 divided by its earnings per share of $20 in the past year equals 14. Do the same for NatWest and out pops the number nine.
Sure, price/earnings ratios have problems galore. So it’s amazing how often they produce a vaguely sensible number. I wouldn’t use PE ratios to compare stocks between sectors or regions. But to sense-check a single share price, it’s fine.
Having said that, some tweaks can improve the predictive power of PE ratios considerably. One of their many failings is that earnings volatility fools them. Another biggie is they are blind to how much debt a company has.
The first can partly be solved by stripping out so-called extraordinary items from the earnings per share number in the denominator. This helps to make sure that one-offs, such as gains on disposals, legal settlements or writedowns, don’t flatter or damp your valuation.
Alternatively, perhaps the underlying business is just lumpy — as for makers of huge machines, say. Here it can pay to take an average over a few years to smooth fluctuations in EPS.
On the problem of leverage, you can replace the numerator in a PE ratio with enterprise value, which includes a company’s debt. Meanwhile, the interest as well as paying down of this debt (amortisation) is added back to earnings.
Using enterprise value to earnings before interest, taxes and amortisation (EV/ebita) goes some way to resolving the variation in capital structures. This is common between industries (those spewing cash such as supermarkets can take on more debt than project-based ones) but occurs within sectors too.
Take the UK energy companies Shell and BP, which are in the news this week, denying that the former is eyeing up the latter. They both have exactly the same PE ratio of 10 times if you use consensus earnings for this year. But BP has more than twice the debt relative to equity as Shell and its EV/ebita is double. Not as cheap as it seems in other words.
Finally, I would recommend that readers save a very basic discounted cash flow model on their computers. Not only are they another tried and tested way to help pick stocks, you can also use them to value almost anything with an income stream — a rental property for example.
I have to talk to our legal department, which will probably add a hundred pages of small print, but please email me at the address below if you would like me to send you a very basic Excel model that only requires a handful of inputs to produce a company valuation.
The idea behind most discounted cash flow methodologies is first getting from sales and profits to free cash flow. The latter is the purest form of cash a company produces, taking also capital spending and tax into account. Estimate this as far out as you can. Five to 10 years is fine.
Only two more inputs are needed. Because time stretches beyond the model to infinity, you need a “terminal growth rate” to derive a “terminal value”. It’s basically a huge dollop of value that represents the future.
Unless there is a bloody good reason, I always use nominal GDP as a proxy for the terminal rate — call it 2.5 per cent. Finally, you need to convert those future cash flows into today’s money — a so-called “present value”. This requires a discount rate.
Think of a discount rate as the interest you must pay for the fact that most of a company’s cash flows don’t come to you immediately. So it makes sense it would be higher the risker a business is. This is an oversimplification, of course. Just look up an average for the sector and use that.
PEs, modified PEs, and a basic discounted cash flow model. Pretty much all most retail investors will ever need. Happy stockpicking!
The author is a former portfolio manager. Email: stuart.kirk@ft.com; X: @stuartkirk__
