There is not a colossal amount of news today. Positioning continues ahead of next month’s Italian election, with Monti and Berlusconi both making media appearances. Berlusconi is in talks with the Northern League and, according to Bloomberg, may be close to an agreement that could hurt the left-leaning centrists backing Monti. In the US, negotiating positions are being staked out ahead of the debt-ceiling debate. McConnell has said that the issue of tax increases has been dealt with and is now off the table, while Pelosi has said that not enough revenue was raised by the fiscal cliff agreement.
On the subject of the debt ceiling, I said in a note last week that I did not think that the platinum coin option was realistic, because it would run the risk of making Obama look like the irresponsible party. On the issue at hand, any strength he has in the negotiations comes from his ability to pain the Republicans as the irresponsible party, and the platinum-coin option could paint Obama in that light; and in the longer term, Obama needs to maintain an aura of discipline and probity. Another option that seems to be off the table is for Obama to declare the debt ceiling unconstitutional on the basis of the 14th Amendment, which says that the validity of the debt of the US “shall not be questioned”. The White House has explicitly ruled this out. I continue to think that the most likely outcome is the one I sketched in last week’s note: some reduction in spending tending towards Obama’s end of things, modification of the sequester and a raising but not elimination of the debt ceiling.
I am once again thinking about tools. If macro investing is, as I believe, all about having the right model, then the more modelling tools one has at one’s disposal, the better. This is one of the most important dimensions along which personal improvement is possible for the macro investor. I have a few specific ones in mind:
– Better statistical knowledge. I am taking an online course in statistics this quarter.
– More economic data synthesis packs. The pack for the US is almost finished, and I would like to produce similar things for other major economies.
– What about the idea of a data mine? I tend to make a model starting from first principles and then see whether it is any good; but there is something to be said for knowing about apparent relationships between different data series, and for being able to search for correlations when a theoretical approach has not got you anywhere. My idea is to have a spreadsheet containing a large number of data series, including lags and rates of change, into which one could put a given series in order quickly to find its strong relationships.7
– What about neural networks? I have never known much about these, but I am working my way through Scott Armstrong’s “Principles of Forecasting” which contains an enticing short paper on the subject. The implementation actually looks pretty simple and I wonder whether a neural-network approach might be better for GDP nowcasting than the composite-index approach of my data packs. There might even be package for this kind of analysis in R. I am going to look into it.
I am inclined to be short equities. My equity model suggests that they do not have much further to rise unless a) credit spreads fall below post-crisis lows into Great Moderation territory or b) earnings increase. The former is possible but unlikely; the world changed in 2008 and credit spreads have not been able to get narrower than they are at present since that change. The latter is unlikely because of fiscal austerity in the US. Although austerity will happen in a mild form, a reduction of the government’s deficit ought to reduce the private-sector surplus. 12m trailing earnings on the S&P 500 have been flat for a year. So I do not think it would be unreasonable to be short. If earnings increased or credit spreads narrowed through their apparent floor, that would be cause to reassess the trade.