The “experts” may yet be right

In 2016 I warned, apparently wrongly, that financial markets would be driven mainly by political shocks rather than by monetary policy or economic data. This year I have decided to double down and claim that politics will be even more dominant in 2017. I justify my obstinacy on two grounds:

- First, it is not quite true that markets shrugged off the impact of the Brexit vote and Donald Trump’s election. Both these political shocks had huge financial effects. They just happened to be confined to bond and currency markets, while leaving equities relatively unperturbed. The Brexit vote caused the one of the biggest currency devaluations in recent British history (whether this devaluation will exceed 2008-09 remains to be seen). Trump caused the biggest one-month loss ever recorded for holders of US treasury bonds. If similar shocks lie ahead this year in France or Italy, European bond investors should brace themselves for even bigger upheavals.

- My second and more important reason for believing that 2017 will be an even more political year than 2016 for financial markets is linked to Tsar Nicholas’s tragic remark. The fact is that neither Brexit nor Trump have yet happened. All that we had in 2016 was an announcement effect. The real significance of these upheavals has not yet permeated market thinking, just as the real significance of the Great War had not yet struck Tsar Nicholas in January 1917. Britain’s commercial relationship with Europe is the same today as it was before the referendum, and the Trump presidency is just 10 days old. Those of us who predicted huge market dislocations can therefore still claim that we weren’t wrong, only premature.

This might sound like a lame excuse, but it draws attention to something important—a profound change in the way financial markets now operate.

Since 2008 I have often discussed with clients how financial markets seem to have lost the predictive powers described in economic textbooks. Instead of predicting the future, markets have become reactive, driven mainly by backward-looking economic data and corporate results.

This change in financial behaviour explains, at least in part, why so many economic “experts” (myself included) were so wrong about the immediate market reactions to Brexit and Trump. We believed, as described in the textbooks, that investors, businesses and consumers would rapidly change their behaviour on the basis of changing expectations about future policies, even though those policies would take years to implement.

For example I expected, along with most conventional economists, that banks and businesses in Britain would change their investment and hiring plans well in advance of any actual changes in trading and regulatory regimes. I also thought that consumers and homeowners would anticipate these business cutbacks and would immediately respond by spending less and saving more.

The obvious reason why businesses and consumers did not in fact tighten their belts in this textbook fashion, is that many did not believe the “expert” predictions that Brexit would affect trading conditions in any significant way. This skepticism about expert predictions was partly justified by the experience of the 2008 crisis. It has now been greatly amplified by the unexpected reaction of financial markets both to Brexit and to Trump, as Louis pointed out last week.

But the unexpected indifference to policy shocks among investors also has another explanation that comes up repeatedly in conversations with active macro-traders. A combination of regulatory and technological changes has shifted the balance of power in even the most liquid financial markets against investors who use economic or political analysis to try and anticipate future policy changes. Instead the balance has moved in favor of algorithmic or passive trading strategies based on sophisticated analysis of correlations and trends that have worked well in the past.

This is not just a point about short-termism due to daily liquidity, or monthly performance reporting, or regulatory reforms that have reduced liquidity and squeezed proprietary trading. The more interesting issue is the deeper tension between active investment management, and passive or algorithmic trading. Active investment management is predictive because it tries to profit by anticipating future changes in economic or political fundamentals. In contrast passive or algorithmic trading is reactive because it tries to profit by analysing trends, correlations and anomalies that have worked well in the past.

Algorithms that are based on past market behaviour or economic correlations cannot, by definition, anticipate structural regime changes of the kind that could result from Trump’s protectionism or Britain’s decision to shift from a service-led to a manufacturing economy. The more markets are dominated by passive trading that reacts to surprises in backward-looking economic and corporate data, the harder it becomes for forward-looking investors to hold large positions that try to anticipate policy changes whose impact may not show up in economic data and corporate results for months or even years. And the more these forward-looking investors are overwhelmed by passive strategies and algorithmic trading, the more reactive, as opposed to predictive, markets become.

These caveats and excuses notwithstanding, I firmly believe that markets will, in the end, respond to policy shocks. Most of the potential responses to the current policy upheavals were eloquently described last week by Louis. In my next paper, to be published later this week, I shall look briefly at the key points of difference between our views.