Why I Need Augmented Decision Making
Photo courtesy of Mr KC Kan, Kurashiki, Japan.
This post was first published in http://chinesevagabond.blogspot.sg/, who is not to be mistaken for a MadChinaMan but one who would be too free-spirited to be bound by the confines of island mentality.
It is a stroke of good luck that he found the mundane topic of trading and investing mistakes amusing enough to expound upon, bringing in the Phd ammo for our enlightenment.
|Fig 1. Property price index from http://www.tradingeconomics.com/|
Gnashing and wailing, seasoned with uncertainty about the property market went well with soft boiled eggs. One moment, two well-defined eggs; the next moment, sticky fingers reached for the black sauce bottle, and spoons or fingernails were used to remove bits of eggshell. Extended slurping and soon, both eggs were gone, and talk about the property markets resumed.
A change from over seven years ago in Singapore was a large number of retirees with money. Or at least speaking as if they had money. There was a false positive here, because the poor elderly ate at home, if at all. For the frugal sixty-somethings, who benefitted from the Confucian work ethic and had invested in property from the 60s to the 00s, the conversation inevitably drifted to: “wah, rent renewed, so much lower!” or “Don’t talk so loud, my flat is vacant!” or “My apartment used to be worth $1.4M, now only $1.2M!”
I listened with envy. I trade, invest, and shop badly. I am aware of my foibles and do not know how to overcome them. Perhaps “artificial intelligence” may rescue me from my stupidity. Such weaknesses included:
- Apophenia (wiki see here), or the inclination to detect patterns in randomness. Was that a face I saw in the clouds? Was Jesus in my toaster? Or in Fig 1 above, was there a trend?
- I do not take careful notes of the situation at any point in time; I do not make explicit my assumptions for decision making or analysis; I do not review past decisions with any precision. An example of the above is my post in 2010 on: Why I think Sky Terrace BTO will be worth $1.4M. Perhaps at some time it was? Did I check? A resounding no.
- Under and overreaction to news was well documented in research. Under reaction was where say there was news, and instead of that news being incorporated into the price fully and immediately, the effect of the “data” was spread out over a longer period. In other words, “stale” news could predict. Overreaction was the opposite. Perhaps such under and overreaction cancelled themselves out, so that incorporation was prompt? Alas, that was not so. Over a short term, prices underreacted to news, but over a longer period, prices overreacted to news. Such historical under and overreaction are the aggregate actions of human beings, and likely to be stable and a source of returns, until machines take over in the future. The explanation for under and over reaction lie in psychology. Barberis, Shleifer, and Vishny (2005) referred to conservativism or belief revision (wiki see here) and the representativeness heuristic or seeing the similarity (wiki see here) to explain underreaction and overreaction to the markets. I might suffer less in these areas than others, which would explain why I do not let profits extend with the trend. Being slightly different is inadequate; being aware of self and others and how that awareness could be used was much more important.
- Finally, I did not know what news was important. By definition, not all news was relevant.
Was it time to buy property in Singapore? The orange line in Fig 1. above appeared to be a good fit, but what was the evidence to project a future upward trend? On hindsight, the run-up post-1992 can be attributed to HDB upgrading feeding the private market; post-1997 fall from the Asian crisis; flatline in the 2000s from SARS and whatever else I had forgotten; and post-2008 is the global, or in the case of Singapore, the China liquidity gusher.
The post 2008 torrent collided with the general elections of 2011 and 2015. In my recent two month trip to Singapore, the number of foreigners had visibly and appreciably shrunk. My look-at-me-I-am-so-important-and-rich-unlike-you-people-your-government-hates detector was mostly dormant. The relaxation of measures reported on 11 Mar 2017 (see here) was recognition of the current slump.
To go to the other extreme was the foreign exchange market. The argument that prices incorporate usable information was more compelling, as the market was liquid. If there was any market in Asia deep enough for machine trading to capitalise on human beings, that was it. Many of my friends make a fine living trading for themselves, so they have intuited how to be successful. Fortunately, they need not face the machines in a big way. The next generation of retirees may no longer be talking about the property market, but about “my algorithm which I cannot tell you because it works best when few people know but I have a special offer for you anyway.”
I am glad I will be dead, when machines take over. Until then, I still need computer support. Damned if you do, damned if you don’t.
Barberis, N., Shleifer, A., and Vishny, R.W. (2005). A model of investor sentiment. In R. Thaler (Ed). Advances in Behavioural Finance. Princeton University Press, Princeton.