Donald Rumsfeld, the former US Secretary of Defence, attracted ridicule for his haiku-like comments about the limits of knowledge:
There are known knowns; there are things we know that we know. There are known unknowns; that is to say there are things that, we now know we don’t know. But there are also unknown unknowns – there are things we do not know we don’t know.
The reason for the derision had much to do with his belligerent brand of politics and little to do with the substance of what he was saying. Because Rumsfeld was making an important point. When it comes to human affairs, including what happens in economies and financial markets, it is the unknown that has the greatest effect.
It has implications for economic thinking. There are many studies that show that economic forecasts have a poorer success rate than tossing a coin. They tend to work when they are obvious and are rarely accurate when something surprising occurs. Friedrich Hayek once said that “the curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.” He probably didn’t apply that point to economic models, including his own, but it is especially true of their inadequacies.
There are a number of obvious reasons why this is so. One is that economic forecasts tend to be a projection from the past, and so seminal changes tend not to be detected. Second is the illusion that economics can be a science. The method of science is to aggregate data or apply mathematics in order to identify what is unchanging, invariant. That means projecting from the past. It quickly results in statements of the obvious or circular arguments when applied to economics, and they do not lead to good predictions.
A third problem, particularly pronounced in the meta money markets that now dominate global finance, is that the mainstream economic and financial theories have become part of traders’ strategies. No theory can include within itself the manipulation of the theory by human beings. Especially when the whole thing is computerised as, is the case with high frequency trading.
The metaphor du jour for the limits of finance knowledge is “black swan events”. This is not even a good metaphor (I was playing a game of golf a few years ago with Houses and Holes and, as I stood on the tee preparing to play a predictably poor drive, I glanced across and saw four black swans staring at me — they are far from scarce). The implication of the metaphor is plain wrong: that economic forecasts are right most of the time except when a “black swan” comes along. Economic forecasts are wrong most of the time and this is entirely to be expected.
Why is this important? Because no-one ever made a buck saying they didn’t know something. Yet knowing the limits to knowledge is essential to intellectual rigour. It is not just the question of falsifiability that Karl Popper identified (that a hypothesis can only be potentially true if it is possible to show that it is false). It is s question of simple honesty. The management thinker Peter Drucker, a sharp critic of economic theory, once said, nostalgically, that economics used to be a humble discipline in which practitioners admitted they don’t know much. That is certainly no longer the case and the reason is simple. There is no money in admitting lack of knowledge. Ergo, don’t admit it.
The same applies to financial advisers. To say that they don’t know something is far from reassuring to clients, so they do not do it of course. But the supposedly superior knowledge of “professional” fund managers in Australia has done little in terms of outperforming the market. As the ABC reported a couple of years ago:
the 10-year rolling average return, the median super fund in a balanced fund where 80 per cent of Australians have their money – was just 4.77 per cent a year – less than five per cent a year. In nominal terms before tax, your money would’ve been better off in the bank in a term deposit, or invested in a government bond.
Those returns wouldn’t have improved much given the weakness of the markets since 2009. But you won’t see many fund managers admitting that they really didn’t know much about the markets. Unless, that is, they are thinking of a career change.
The same inability to admit the limits of knowledge is endemic in the climate debate. In this excellent article in the AFR, Garth Paltridge nicely documents how science is being corrupted by false certitude in the global warming debate:
Attempts to resolve the arguments are plagued with problems, a lot of which are inherently insoluble. There are many aspects of the behaviour of the natural climate system and of human society that are unpredictable in principle, let alone in practice. But perhaps the biggest of the underlying problems, and it is common to both arguments since it inevitably exists when there is large unpredictability and uncertainty, is the presence of strong forces encouraging public overstatement and a belief in worst-case scenarios.
It is especially a problem when applied to risk and risk management. This creates the illusion that the unknown can, within limits, be anticipated and priced. As a fund manager writing in The Economist noted just after the GFC, many money managers thought they were covered against shocks by their use of risk defrayal mechanisms, only to get a nasty surprise. To use Rumsfeld’s metric, risk only applies to known unknowns, not unknown unknowns. As a risk anthropologist friend of mine notes, what is needed is a way of dealing with “vulnerability” rather than risk. The implications of this for banks are pretty obvious.
The Harvard Business Review had an article about nescience and the need to have a “chief ignorance officer” in business to be more effective. That is not a bad idea in economics, finance, risk management and even science. It at least opens up the possibility for more rigour. And, like Popper’s negative test of falsifiability, it can create the prospect of knowing when you really do know something. Only those who freely embrace ignorance have much prospect of reasonably often being right.