November 22, 2018

· website

This post was co-authored with Andy Eggers (Nuffield College, Oxford)

Will Theresa May’s Brexit deal win the support of Parliament? A view is emerging in Westminster that May’s EU withdrawal plan will succeed in the same way that TARP (the Troubled Asset Relief Program) cleared the U.S. Congress in September of 2008: an initial rejection is followed by a stock market crash, which concentrates legislators’ minds and clears the way for successful passage.1 A market crash may indeed be just the thing some MPs need to secure their support for May’s withdrawal bill, whether because it convinces them of the bill’s merits or provides the cover they need to take a politically unpopular position.

The question is whether the market would play along with this plan. The fundamental problem is that the more transparent the “TARP model” for passing the EU withdrawal bill becomes, the less likely it is to work.

Suppose that, having seen multiple articles in the FT about the “TARP model”, market traders become certain that the initial attempt to pass the bill will fail. Then the initial failure of the bill provides no new information. MPs who hope for political cover from the markets will be disappointed, because all of the expected economic damage resulting from an initial rejection of the bill will already be “priced in”.

This logic can be taken one perverse step further. Suppose now that those FT articles on the “TARP model” leave it uncertain whether the bill will succeed in a first vote, but they convince traders that the bill will pass in a second vote if (and only if) the initial failure precipitates a market drop. An initial failure happens. A naive trader might sell at a discount, contributing to a market drop that convinces MPs to pass the bill in a second vote. But this naive trader loses money when the bill passes. More sophisticated traders refuse to sell at a discount, which keeps prices high and prevents a market drop. The paradox, then, is that the belief that a market drop will lead to bill passage prevents a market drop from happening, which prevents bill passage.2

The upshot is that the TARP model can work as a political strategy for Theresa May only if markets remain uncertain about whether it can work as a political strategy for Theresa May. If May knows that she will lose an initial vote but win a second vote when the first vote triggers market mayhem, and the market knows this too (or knows that she wouldn’t go ahead if she didn’t know this), then there may not be market mayhem and the plan won’t work.3

For predicting political outcomes, the first-order task is usually to determine what political actors intend to do; typically, the more uncertain those intentions, the more uncertain the outcome. With all the recent talk about the “TARP model”, we may be in an unusual situation in which the less uncertain the intentions of political actors, the more uncertain the outcome. As long as markets remain unsure of how MPs will respond to market events, market events may guide political developments according to the TARP model. But if MPs’ intentions become too transparent, all bets are off.


  1. In 2008, during the initial stages of the financial crisis, the enabling act for TARP was voted down in the US House of Representatives on 29 September 2008 by a 205-228 margin, precipitating a 8.8% fall in the S&P 500 that day. After the Senate voted through a revised version 74-25 on October 1, the revised version was voted through by the House 263-171 on October 3. The conventional wisdom regarding what happened between the two House votes is not that the minor changes to the legislation won over 60 representatives who had opposed the initial version, but rather that the stark market reaction did the job. [return]
  2. This argument can be made in the form of a game theoretic model. Indeed we have been in the early stage of working on a paper describing this dynamic between markets and politicians, but events move more swiftly than academic research, and thus this post. The main conclusion, which is robust across several ways you might describe the market-politician interaction, is that markets cannot prevent politicians from making economically costly policy decisions, at least not 100% of the time. Because markets are forecasting the future economic conditions that will actually occur, after the politicians make their decisions, the market cannot send a signal that is always successful at preventing politicians from making the economically costly decision. If the market decline always prevents the costly political action, the market should not have declined in the first place. It is possible to write down models where markets sends more ambiguous signals that politicians respond to only some of the time, but it has to be an uncertain enterprise in order for the signalling to work. [return]
  3. What was different in 2008 that made it possible for the market to successfully signal to politicians about the consequences of inaction? No one had given the game away. There was genuine uncertainty about the first House vote before it happened and there was also genuine uncertainty about whether the politicians would change their minds in response to market action. [return]
All Posts by Date · All Posts by Category