Jefferson V. DeAngelis

Jefferson V. DeAngelis


May 29, 2019

In the classic comedy movie Airplane, one of the more memorable scenes is when the characters engage the autopilot to keep their plane airborne. The catch, however, was that the autopilot could not land the plane. Ultimately, the passengers had to rely on ex-fighter pilot Ted Striker to overcome his (comical) pathological fear of flying to land them safely. While unintentional, the movie strikes a prophetic note; that humans should not be so quick to turn over the reins to automation.

Autonomous cars and airplanes hold promise. Even in financial markets, the move is well underway to greater automation, but it is premature to completely remove the human pilot. Unfortunately, we have seen instances both for vehicles and airplanes where reliance on automation can have disastrous results. While artificial intelligence means that computers can learn, it also implies there will be mistakes along the way. How computers respond to input signals can differ from humans, not always for the better.

The markets incorrectly assumed that the Fed’s decision to flip on the autopilot switch meant that it was safe to assume risk. Autonomous trading systems like the systems found in modern airplanes and automobiles are fraught with risk. Excessive risk-taking is the result of over-confidence around a certain outcome, ignoring the fact that there is a distribution of outcomes, often with nearly equal probability of bad outcomes. Such confidence breeds complacency and crowded trades.

It was apparent at the end of last year that policymakers were headed in the wrong direction by draining liquidity from the system. The markets were signaling that the Fed was making a mistake and President Trump was quick to remind them. Fortunately, they averted a more significant meltdown by clearly communicating to the financial community their intention to pause. The financial markets were quick to respond. Stock returns during the first quarter hit historic highs.

Investors may have overreacted by building into their forecasts a reduction in interest rates and accommodation by Fed Chairman Powell. Clearly, Powell is now more closely aligned to the thinking of his predecessors, willing to risk the economy running too hot and fearing that it may be catching a cold. Then again, he is unlikely to lower interest rates prior to a 20% correction in stock markets. At current market levels, he is stuck in a holding pattern.

Through the first four months of the year, interest rates fell rather significantly, and stock prices surged. Additionally, there were several other factors stoking investors’ confidence. Chinese policymakers embarked on another round of stimulus and both the Chinese and United States trade negotiators were signaling progress on trade negotiations.

Algorithmic trading routines are quick to identify signals but can miss the nuances of negotiations. Signals coming out of the U.S.-China negotiations seemed positive right until their collapse. The signals belied the challenges of trade reform given cultural differences of Chinese collectivism and American individualism. For the Chinese, an idea is the property of the state and innovation is possible only if it benefits the state. Meanwhile, the U.S. government seeks to protect intellectual property, to allow individuals and companies to profit from the development of their ideas. It is much easier to arrive at numerical quotas for trade than to bridge cultural differences. When the stakes were t-shirts and soybeans it was possible. When it involves technology and intellectual property, it may be impossible.

With trade talks having broken down, markets need to adjust expectations for a more protracted and entrenched fight. The recent escalation of tit-for-tat tariffs is a sign that both sides are confident in their ability to manage the political and economic fallout, at least in the near term. While Trump offered some positive soundbites of late, the timetable has clearly slipped to summer or fall before any resolution.

The standoff indicates that the comprehensive reforms sought by the U.S. might remain elusive in the end. The Chinese are less willing to yield to attempts to codify protections of intellectual property rights or abandon their “Made in China 2025” project. Still, the Chinese appear willing to concede the need for a more balanced trade relationship and are prepared to buy more U.S. goods. Trump will eventually settle for a lower trade deficit and the appearance of a victory.

Against this uncertain trade backdrop, global growth continues to slow. Economists have erred in their understanding of the impact of the trade war on supply chains and business planning. They have focused too simply on the impact on GDP. Decisions on the margin by businesses to delay investment or establish new supply chains are having a more pronounced impact on global growth. Inventories continue to build, and traditional relationships between inflation, growth and labor markets have broken down. Central bankers continue to push all the buttons and switches without the expected “normal” responses. Their actions have become more muted. They must react with more unconventional measures. Or, so they think.

The economic ship is not responding to new directions precisely because it is laden with the increasing debt levels which central bank policies promoted. Inflation would be welcome because it could potentially reduce the economic drag of debt service. Rather, the risk of deflation looms larger. A scarcity of income to service debt is equally as likely as a devalued dollar.

The “new” normal of our low interest rate environment has become counterproductive. Promoting policies that lower interest rates incentivize bad behavior via excessive risk-taking. We have made several mistakes in monetary and fiscal policy that have led to the potential for credit problems. These credit problems become more acute should we fail at trade negotiations and enter a recession. To a person of reason, an adjustment seems logical. But it requires courage. If we continue to act as if we are locked in a do…loop within a machine learning algorithm we are in for a hard landing.


To download the PDF version, click here: 2019-05 Autopilot


©2019 Northwest Passage Capital Advisors LLC. All rights reserved.

2019-05-29T09:21:15+00:00 By |