The China Syndrome

Shanghai Stock Exchange Composite Index

The recent stock market sell off owes little if anything to fundamentals, and everything to herd mentality. Nothing we found out when the sell off began was anything we didn't already know. China's economy is slowing, energy supply still exceeds demand and is US interest rates are likely to rise towards the end of the year. We've known all this for months, so why the panic in US markets?

It was a knee-jerk reaction to a sell-off in China. That in turn was a bursting of an equity bubble that had been created after the run-up in property prices stalled and many of China's new-to-the-market fairly naïve investors were looking for the next big opportunity. Many were over-extended and had to sell as market began to turn down, turning a gentle slide into a route. But that was always going to happen - PE ratios were absurdly high and prices had become completely disconnected from the underlying fundamentals. (Even now,they are probably too high - the market is 50% above where it was 18 months ago).

But instead of looking at China's melt-down as just a much needed market correction and not an indicator of any real change to the economy, investors in Europe and the US panicked and rushed for the exits too.

That may be over-stating the case. There are several ways in which the "China Syndrome" might impact US companies. First the Chinese government's response (lowering interest rates and devaluing the Yuan) were hurried and cast doubt on the Party's ability to manage the market economy it is creating. Second, the devaluation and the denting in consumer confidence among China's neauveau riche will hurt consumer technology and luxury goods sales; much of Apple's recent profit growth has come from China.

But the scale of the investors' reaction in the US vastly outweighs the impact of a likely 1% reduction in China's GDP this year (which we already knews was on the cards). If US companies relied exclusively on China for sales and profit, then the drop in US stocks would be in the right ball park - but US exports to China ($123bn in 2014) account for less than 1% of US GDP ($17.4tr). Granted other countries to which the US exports might also themselves export to China so that the effect is larger than just the US' trade with China; but even so the reaction seems quite out of proportion with any reasonable estimate of the impact of  China's slow-down on the US economy.       
    
All of which supports my contention, an observation I made to a finance colleague 20 years ago, that finance is less about pricing of assets and risk than it is a function of rumour, social network and herd behaviour. If expectations of the present value of future cash flows determined asset prices none of last week ups and downs would have happened.  

Collective intelligence

One distinguishing feature of human civilization is out ability to share information with each other and pass accumulated information across generations. Standing on the shoulders of giants has enabled the development of a huge body of knowledge and understanding that enables us to manipulate our environment in ways that no other species has achieved.

Information sharing, while central to our species' success, is nevertheless fairly crude and imperfect for two reasons. First language requires that knowledge be encoded and that encoding leads to a loss of fidelity in transmission. This is the flip side of tacit knowledge. Second, some people decide that sharing reduces their relative advantage on which resource appropriation depends. In other words keeping information to oneself and exploiting it leads to higher rewards than sharing it.

Recently there have been some luminaries in the IT world who have expressed the fear that machine intelligence will soon surpass human intelligence and at that point all bets are off.

My take is slightly different. It's not the power of single computers that presents a threat but their collective intelligence; computers have no qualms about sharing data, and data can be communicated losslessly between them. So even if they are individually less capable at the nodes, their ability to access collective information though the network and develop a collective intelligence (akin to "organizational intelligence" - see Jim March's writing) vastly outstrips ours. That is what I think should give is pause.

That leads to the question of societal organization; Keynes imagined a world 100 years from the time of his writing in 1930 in which machines did almost all the work and people lived lives of leisure. Yet that looks unlikely; more people are working and working at least as long as they did 100 years ago - his anticipated life of leisure seems a cruel joke.

If indeed machines do soon acquire the ability to performs not just repetitive manual task but a wide range of more complex cognitive tasks, the way we currently organize society looks untenable. 80% unemployment with the owners of capital benefiting from the machines' labors is a recipe for high social unrest and instability. So we may have to grapple with the issue of some transfer of wealth to provide for those who have been displaced - otherwise, we risk a society that looks rather like the middle ages with a handful of unimaginably wealthy barons enjoying the services of a huge number of serfs, living in abject poverty.

Some might say that with 45 million Americans already living on or below the poverty line, we are well on the way.