At the top of yesterday’s note I wrote: “Today’s central bank action is the big news for the market. Did anyone think central banks would fail to provide swap lines to each other in the event of a credit crunch?” In other words, I did not think the action was that significant. Why, then, did the market rally? The best explanation is provided by Andrew Hunt: the authorities have successfully signalled that they will act to prevent another Lehman-type event. Given their actions after the Lehman collapse, it always seemed likely that that was the case; but market participants had started to worry about a credit crunch, and perhaps a small probability of such an event was being priced into risk assets. Now that small probability has got smaller, and markets have rallied. It is possible to make a simple model of how this can happen. Imagine an index at 100. Risks to the index are evenly balanced, with a 50% chance it will rise by 20% in the coming year and a 50% chance it will fall 20%. Now imagine that the downside risk is split between a 40% chance of a 10% decline, and a 10% chance of a 60% decline (which seems not inconsistent with the idea of Lehman II). If some action is taken that makes no difference to the upside, but changes the downside probabilities, the index level can rise. To get a 4% rally in the index, the probability of the extreme downside scenario has to fall from 10% to 2% — i.e. from pretty unlikely to highly unlikely. This is what I think happened yesterday. If yesterday’s action precipitates a sustained rally, I will have been wrong.

Whenever there is a reasonably big move in the markets, people get excited about just how big it was. I am completely uninterested in these discussions. I do not think that individual days tell you anything at all about what the market is going to do next. The S&P 500 yesterday moved 4.2%, or 2.04x its average true range of the previous 10 days — the latter is fairly uncommon, in that over the past 20 years, it has happened on around 1% of all trading days — so perhaps people think it is unusually predictive. This morning I have tested the predictive power of strong up days over the past 20 years in the S&P 500. My strategies buy the market on a given day and hold for two weeks. The metrics used to evaluate them are the percentage of profitable trades, and the profit factor (=profits/losses). Here are the strategies:
  • Buy at random. Percent profitable: 59%; profit factor: 1.15.
  • Buy on up days when market closes in the top 1/3 of the day’s range and true range > average of the previous 10 days x 2. Percent profitable: 55%; profit factor: 0.94.
  • Buy on up days when market closes in the top 1/3 of the day’s range and the day’s move is over 4%. Percent profitable: 55%; profit factor: 0.83.
Unsurprisingly, strong up days are not predictive of good performance in the next 10 trading days. On the contrary, buying on strong up days is a losing strategy even in a market that has generally gone up over 20 years. That makes sense: buying such days will systematically get you in at a higher price than buying at random, and, other things being equal, getting in at a higher price is a bad idea.
Is it time to bet on Egypt? The elections have happened and early indications are a win for the proxy of the Muslim Brotherhood, which stood on a pro-market platform. I think that prolonged civil war is unlikely in Egypt — either there will be a peaceful transition to democracy and the country will be governed by the Brotherhood, or the generals will manage to keep control. Either way, the environment should not be unfriendly to business. I wonder whether there is a revolutionary risk premium in Egyptian stocks that might present a buying opportunity.
Bloomberg quotes a fund manager as saying the economy is in “A classic Irving Fisher debt deflation.” I do not see how anyone could say that. Fisher’s debt-deflation story involves a vicious cycle between debt liquidation and falling prices. The fall in prices is crucial. Fisher himself says that “When over-indebtedness stands alone… the resulting ‘cycle’ will be far milder.” Happily today, unlike in the Great Depression, we have both fiat currencies and central banks that are ready and willing to use the freedom they provide to prevent deflation — exactly as Fisher prescribed. There is a problem of a debt overhang, but the analysis that is useful here is not “classic debt deflation” but classic circular flow. This model breaks the economy into households, firms, government and the external sector, and envisages a simple flow of funds between them. It suggests that for one sector to run a financial surplus — such as households that want to pay down debt — at least one other sector must run a deficit. At present, the government is helpfully running one. (This model has short-term implications for austerity programmes — if government deficits are cut, household surpluses will have to fall). 
Ever since I have started trading for myself, I have asked myself the question: how is it possible to understand the market? It occurs to me that human beings are capable of focusing on only one or two elements of a situation. This means that the most important discretionary input it is possible to have in an investment process is to pick out the salient elements, and that is the skill I have been trying to learn is not so much comprehension as discrimination. What can one do, then, when it is necessary to assimilate several different data points to reach a conclusion, something at which human beings are very poor? Daniel Kahneman suggests some kind of algorithm. For example, when evaluating a fund manager one might score him on a number of different areas and combine the scores using a set formula. Recently I have been using a different approach to understanding the economy: leading indicators. I have created a leading economic indicator and pretty-much replicated ECRI’s leading index for inflation. I thought yesterday that it might be worthwhile to create composite indices for other variables that are affected by a range of different factors, and I am going to experiment with so doing for short-term moves in the equity market.
Data:
China cut reserve requirements 50bps (which means a fall from 21.5% to 20% for the largest banks).
ADP non-farm employment change 206k b.e. by a large margin.
Chicago PMI 62.6 b.e. All the regional surveys are in and suggest a small rise in the ISM PMI from 50.8 to around 52.3.
Pending home sales 10.4% b.e. in a big way.
Australia AIG manufacturing index 47.8 — 9th month below 50 in 2011.
Australia building approvals had another big fall.
Australia RS 0.2% d.e.
China PMI 49, d.e. and fell below 50. HSBC China PMI was revised down from the flash estimate of 48 to 47.7.
Eurozone PMI was as the flash estimate, 46.4.
UK PMI 47.6 b.e. but fell from last month.
Next 24 Hours:
Initial claims
ISM PMI
UK construction PMI
Non-farm payrolls
Advertisements