Project simplification

For the first year of the Thrifty 30, the way I discovered companies remained unchanged. I screened the market for bargain stocks (share price lower than the value of current assets minus all liabilities), thrifty stocks (share price low relative to average earnings in companies that seemed financially strong), and Magic Formula stocks (highly profitable companies at cheap prices).

In 2011 though, I’ve fiddled incessantly with the screens as part of what I might have dubbed ‘project simplification’, a behind-the-scenes effort to reduce the amount of data I collect and analyse, with the intention of reducing the complexity of my spreadsheets.

You can see the new screens, and explanations of how I constructed the new and old screens on the T30 shortlists page.

Spreadsheets are a significant element of my investment process, both in discovering companies and analysing them. They are, therefore, a significant cost because one of the simplest laws of investment is the longer you spend on one company, the fewer companies you can investigate. Since the first step I take in analysing a company is to check the data and calculations used in the screens (from Sharelockholmes.com) against the company’s annual reports, I also needed to ensure that the two spreadsheets, screening and analysis, contain the same reduced set of variables.

Using a reduced set of variables ought to be more efficient, like the RISC (Reduced Instruction Set Computing) processers in the ARM-designed chips of your iPhone are more efficient than CISC (Complex Instruction Set Computing) chips probably residing in your PC. You don’t lose anything but effort by reducing the variables (instructions in my analogy), as long as you can reconstruct any complex variables you need from the ones that remain.

That was the logic of my ‘better long-term PE’ post. Since I can reconstruct a similar (perhaps even better) measure to the long-term PE from a company’s average returns on equity and current price to book value, it isn’t necessary for me to collect earnings per share data any more. If I screen on ROE and PB, then that’s one less data item and calculation I need to check.

As part of this process of rationalisation I realised I’ve never added a Magic Formula stock to the Thrifty 30. I have faith in Greenblatt’s strategy and continue to run a Magic Formula inspired portfolio for Money Observer magazine, but it’s a mechanical investing technique and the Thrifty 30 is a manually invested portfolio. I wouldn’t feel comfortable adding Magic Formula stocks unless I was adding large numbers of them routinely to reduce the risk that some, if not many will be duds.

Experience, other people’s research, and plain old intuition tell me, in the words of Winston Churchill:

The farther backward you can look, the farther forward you are likely to see

Which is why I favour calculating long-term average earnings yields over the Magic Formula’s single year earnings yields.

So I’ve dropped the Magic Formula screen. That doesn’t mean I’ve given up on finding good companies at cheap prices. The Thrifty screen does that by ferreting out high ROE companies on low PBs, and the Hidden Champions screen takes the quest for quality one stage further by insisting that ROE was not just high on average over the last 10 years, but high in every sub-period measured by Sharelockholmes (one, three and five years).

Curiously, two companies sit in both the Thrifty and Hidden Champions lists this month. They are Rio Tinto, the mining monster, and JD Sports Fashion. Rio Tinto is not my kind of company, and I know less about JD Sports Fashion than computer instruction sets (and you know all I know about them now), but its stats look quite remarkable. It’s not only among the cheapest high quality companies, statistically, it’s also one of the highest quality cheap companies!

Narborough Plantations, the top ranked Thrifty company appears to have more equity than assets, which I assume is a glitch in the data. According to its accounts total assets are nearly £11m and shareholders equity is only £200,000 less, so equity is pretty much equal to assets and it probably deserves a place in the table.

I’m also glad I’ll get another opportunity to look at Aga, which nearly became one of the inaugural members of the Thrifty 30. It’s a stock I’ve agonised about many times.

Finally I have a growing backlog of Thrifty 30 companies that have published annual reports and are consequently up for review.

1. French Connection
2. Holders Technology
3. Johnson Service
4. Mallet
5. Quadnetics (done most of the work, just requires a decision)
6. T Clarke

So it’s time to stop fiddling with the guts of spreadsheets and start plugging information into them.