Friday, January 19, 2018

Friday videos: forgotten 1991 UK dance

Nobody remembers Bizarre Inc., but this is pretty historic for its effortless switching between hard techno and Balearic in the same song.

Wednesday, January 17, 2018

Still no news, some theory instead

Did my seminar, it might have been good, I dunno.

No real news on the market front: despite yesterday, people are still in a loading-up-the-truck mode in the US equities. $USD has been dropping precipitously, which you should expect when the US government is generating massive new future deficits with their tax plan when the rest of the world is eating away at their deficits through economic growth. Of course, a drop in $USD is great for US businesses with a lot of foreign sales, and a drop in $USD means stocks have to rise in $USD terms to maintain real value.

Gold should go up because the other commodities have gone up. Although the scuttlebutt is that the rise in the price of oil is just being engineered by the Saudis to get the best possible price for their float of Saudi Aramco, in which case you should expect the price of oil to go back down in a year or so to kick US shale back out of the market.

Have been trying to consolidate my view of stock markets:

1. A stock price doesn't rise as a reward for booked earnings, because investors don't buy stocks to reward companies: they buy stocks in expectation of future appreciation.

2. If prices move because of expectations, then your most important factors in generating an algorithm for "fair price" should be expectation functions, not empirical values from the past (ignoring the idea of temporal correlation in time series, since after all that tends toward zero as t increases).

3. So the market is a mechanism for averaging over the expectation functions of all market participants. Importantly, expectation functions of non-participants are worth nothing (except to try to predict when they may change into participants), and expectation functions of participants are weighted by their size of participation.

4. Then, the job of the investor is to try to predict the inflection points for the expectation functions, not the inflection points for prices - though the latter will influence the former, obviously.

5. So basically, your job is to try to calculate E(E(x)).

Seems like the above is even beyond Shiller, so that's probably a pretty decent theory for now.