Pockets of Predictability in Equities

Following on from my ramblings yesterday, I have been thinking about how it is possible to trade pockets of predictability in time for equities. In theory this should be possible: simply buy stocks that are:

  • Underpinned by fundamental value.
  • Experiencing some positive change.

Putting it this way exposes the obvious problems.

  1. It is very time-consuming to identify stocks that are experiencing positive change. Since it is not possible to follow every stock in the world all the time, the commonly-adopted method is to read around for stories, or to hire analysts to live the stories of a small number of stocks (say, a sectoral basket). These methods are highly inefficient.
  2. Salient change is extremely hard to identify and exploit even if you know a stock well. A fortiori, the identification and exploitation of change that will remain salient in several years’ time is even harder — but that is what we, as long-term investors, would be attempting to do.
  3. Most measures of fundamental value rely on predictions of the future. This means that they are vulnerable to the same objections as the identification of salient change.

You might say that one can develop a set of metrics and principles that at least generate an excess of profits over losses over time, at least for a while. This is, indeed, possible, but it implies the exploitation of a particular pocket in space, not a succession of pockets in time in the markets for different stocks.

Thus, two things are needed: a marker or set of markers that point to the possible existence of a pocket in time; and a method for assessing the future prospects of a company that is robust to errors in the identification of salient change.

I have thought of two potential markers for the possible existence of a pocket in time. Both are based on the idea that stocks ought to be underpinned by fundamental value, which in turn is just an implication of the assumption that markets are coherent with the rest of reality (which anyone who attempts to exploit pockets in time must assume), on a desire to keep assumptions about the future as vague as possible, and on a desire to be conservative in order acknowledge the slipperiness of the concept of fundamental value. They are:

  • Price near or below net tangible book.
  • Price near or below the value of future cash flows, assuming no growth or decline in the business, discounted by the required rate of return.

These are not intended as criteria that a stock must fulfill in order to qualify for investment. Pockets in time exist in all kinds of stocks, not just those that display one of my markers. Nor are they the metrics of a process intended to exploit a pocket in space (indeed, I would be surprised if a systematic investment process based on them made money). Rather, it seems to me that if a stock has one or both of these markers, and there is no good reason why it should fall further, and there are reasons why it could rise, then that should be enough to mean that it should not generally trade lower over the medium term. The market for such a stock would be unusually predictable — i.e. there would be a pocket in time. Crucially, it is possible to screen for stocks that begin to exhibit the markers, something that is not possible for most other changes that create pockets in time. So a method based on these markers should, I hope, make for the efficient identification of genuine predictability.

Some readers may find this discussion of process rather self-indulgent. I need it. I am a “meta” person — I am not happy or effective unless I understand exactly why I am doing something. I also believe that one is more likely to achieve long-term success if one understands the reasons for all of one’s actions.

I would appreciate any criticism and any suggestions for other potential markers.

ECB Press Conference

The ECB held interest rates and did not discuss a cut. There was not a lot of new information in the press conference — Draghi is in wait-and-see mode as the effects of the 3-year LTRO’s ripple through the system, and would not be drawn on the subject of Greek writedowns (as you would expect — why do the journalists even ask?). It is clear that the ECB is concerned about the shortage of collateral, with Draghi mentioning the issue early in his statement. Bank loans will be accepted as collateral in the forthcoming LTRO — a decision that was not unanimous — and a list of eligible collateral was published this afternoon. Draghi said that haircuts on the new forms of collateral would average 33%.

Some of Draghi’s comments were useful in confirming my view of the ECB’s attitudes. He said that, “if one is being bold”, it was possible to see the recent fiscal compact as the first step towards European fiscal integration. That is consistent with my view that the ECB will continue to keep the euro crisis in abeyance as long as politicians continue to move towards an eventual fiscal solution. He also said that TARGET2 imbalances within the Eurosystem were a natural feature of a monetary union and did not imply greater risk. This is consistent with my view that the ECB will continue to fund balance-of-payments deficits in the PIIGS and that TARGET2 imbalances will not be a brake on this.


  • Japan core machinery orders -7.1% MOM d.e. Dec. This is a volatile series; November’s increase was 14.8%.
  • Japan M2 growth 3% YOY d.e. Jan. This is pretty high compared to the past ten years.
  • China CPI 4.5% YOY b.e. Jan. Cue much speculation about the pace of monetary easing.
  • China PPI 0.7% YOY a.e. Jan. The slowest pace since 2009.
  • Japan household confidence rose, b.e., and is regaining its pre-quake levels.
  • UK manufacturing production 1% b.e. Dec. A strong number.
  • UK IP 0.5% b.e. Dec.
  • UK trade deficit narrowed to a relatively narrow level, b.e., Dec.
  • BoE held rates and increased QE by £50bn as expected.
  • ECB held.
  • Initial claims 358k b.e. and fell. Downward trend intact.
  • US wholesale inventories 1% d.e. Dec.

Next 24 Hours

  • China trade balance
  • RBA statement
  • Swiss CPI
  • UK PPI