Diary

Market takes a caning

FTSE is down 4.64%. What's really noteworthy is that every share in the Footsie is down. That must be a pretty rare occurance. The biggest hit were miners. The biggest faller was VED (Vedanta), down 13.1%. In the US, the VIX was up 12.7% at the time of writing, sending it to 42.06. Pretty impressive stuff.

What I think is interesting is to  take a look at some of the stuff that hasn't been affected much, or has actually gone up (belive it or not, some stocks have). I think it is useful, because it may offer important clues as to which stocks are likely to be beaten down as low as they can go, and so may be worth considering from an upside. There are some that Richard Beddard over at Interactive Investor has either talked about, or owns in one of his portfolios. I'm not tracking all of his shares, bear in mind, but I can see that AMR (Amour Group) is down a mere 0.24%, MRX (Metalrax) didn't budge, VLK (Vislink) is actually up 0.8%. UKM (UK Mail Group) isn't one of Richards, but it is held by one of his "regulars", so it was worth mentioning.

Miners and banks took a caning, of course, together with Motley Fool favourite AV (Aviva), and there was some surprises like BATS (Brit Amer Tobacco), down 4.7%.

What other interesting stuff is there? Well CHW (Chime Communications) is down only 0.9% - a share that was bought to my attention by Cautious Bull, which I'd like to talk about sometime. CTN didn't budge - not surprising since there's a guaranteed bid of 85p. Alas - if only that would tumble, there'd be free money. DLAR (De La Rue) remained static - this is a company that I also want to talk about. DPLM (Diploma) is down only 0.2%, which I rated as pretty good quality, on a PER of 11. DNO (Domino Printing Scicences) is actually up 0.1%. I had talked about it on Motely Fool yesterday, explaining that I thought it was a quality company that I thought could deliver above-average growth.

Retailers seem very robust today - perhaps not what one would expect at all. GMG (Game Group) - company that Geoff Gannon talked about recently - it hasn't budged. It has 120m in net cash, and a market cap of 74m. I see today that it announced an RNS for a strategic partnership with OnLive. ALY (Laura Ashley) didn't move, and JD Sports was acutally fractionally up. It's not all consistent though: TCG (Thomas Cook) was down 6.8%, MKS (Marks & Spencers) is down 3.9%, and FCCN (French Connection) was down 3.1%.

IND (IndigoVision) - remember I touched on them yesterday - didn't move. GDL, a spinoff in the oil industry, which I looked at a few weeks and decided it was overvalued, is down only 1.7% - not bad, when you consider that all things resource-related are being clobbered.

MFI shorting

To follow up on shorting, the original sentence taken by Greenblatt's appendix to his second edition of the Little Book is:

If we had actually tried this strategy of buying all of the Group 1 stocks and shorting all of the Group 10 stocks over the last 22 years, wewould have had a tough time of it. Instead of earning 15.4 percent (15.2 percent from our longs in Group 1 and 0.2 percent from our shorts in
Group 10), somewhere in the year 2000 we would have had a little problem. Okay, a big problem. Okay, we would have gone brokeâ€"losing100 percent of our money! No matter how long our time horizon, the number 0 does not compound very well!

Greenblatt versus Graham

Geoff Gannon wrote an interesting post on combining low PEs with low P/E or high ROIC.

I wont repeat his arguments, but I thought I'd add a little comment. It's trivially true that:
P/B = (P/E) * (E/B)
E/B is effectively a return on capital. Let's suppose we fix P/E low. If you want a high E/B in the Greenblatt style, then you are choosing a bit of a mixed P/B. If you go for a Graham approach, then the low P/B range implies a lowest E/B range. So you're kinda assuming a reversion-to-mean.

What's interesting, is that I actually have some stats to suggest which one will work better. Penman, in Financial Statement Analysis and Security Valuation, p 18, actually gives a table of P/B versus P/E for different quintiles. If I concentrate on the lowest P/E quintile, then it turns out that the lowest P/B quintile gives the best result (30.0%pa), against the highest P/B quintile (19.7%). It turns out that that they didn't quite get a monotnoic ordering for the quintiles. From highest to lowest P/B quintiles, the returns were 19.7%, 22.1%, 21.6%, 24.3%, 30.0%. So interestingly, if two men enter, one man leaves, I would be inclined to say that Graham would emerge the winner. -- FINISH HIM --, as they say on Mortal Kombat.

A poster on Geoff's blog notes the following:

If you read "Behavioural Investing" by James Montier you will find a graph where he backtested a pure EBIT/EV approach and a EBIT/EV combined with ROA approach for the US stock market. He found that over the long term focusing on high ROA actually subtracts returns from a pure EBIT/EV screen. However ROA reduces the variability of returns a bit. I bet this is also true for the magic formula. Since both EBIT/EV and ROA are mean reverting you combine tail wind (EBIT/EV) with head wind (ROA) when screening with both.
Therefore if using a pure screening strategy I believe including business quality metrics such as ROIC, ROA, ROE will diminish returns. BUT if you are as good as Buffet and Fisher and know how to figure out if the business has a true moat these will probably work in your favor.