Views from the Floor - The spectre of a US shutdown: How spooked could government sensitive stocks get?

Looking to history for clues as to how stocks sensitive to government spending could perform.

The immediate challenge for newly elected speaker of the US House of Representatives Mike Johnson is to avoid a US government shutdown which, if action is not taken, will begin in mid-November.1 We look to history to see if there might be clues as to how stocks which are sensitive to government spending could perform in this environment.

To do so, we construct a government sensitive basket of stocks by calculating the amount of dollar exposure all US stocks have in terms of outstanding government contracts and retaining those with significant exposures. As one might expect, owing to the nature of government spending, we find that sector concentration is highest in Industrials (35%), Healthcare (20%), and Technology (15%) (as shown in Figure 1).

Figure 1. Sector exposure of stock basket

Problems loading this infographic? - Please click here

Source: Man Numeric.

This basket is likely to be impacted negatively during times of uncertainty relating to the spending ability of the government. On the other hand, events leading to resolutions should be perceived relatively optimistically. We capture these events using historical data on debt limit standoffs.

We perform an analysis of how such a point-in-time basket would have performed during the historical debt limit standoffs. In Figure 2, Day 0 represents the historical debt limit deadlines.2 We show performance 30 days prior to Day 0, given debt limit deadlines are likely to be preceded by negative perception due to uncertainty surrounding the government’s course of action. We see that the basket outperforms the S&P 500 Index less and less leading up to these dates.

Figure 2. Negative perception leading up to debt limit deadlines

Problems loading this infographic? - Please click here

Source: Man Numeric.

In Figure 3, Day 0 represents historical dates on which the government reached a resolution on the debt limit standoff.3 We show performance 30 days following Day 0. On average, the basket of stocks recovers around 1.5% during this period owing to their positive perception and avoidance of cashflow disruptions from contracting activity.

Figure 3. Positive perception after debt deal

Problems loading this infographic? - Please click here

Source: Man Numeric.

We also study the most recent 2023 debt limit stand-off. The US Treasury reached its debt limit on 19 January 2023. However, the projected X-date when the government would not be able to fulfil its monetary obligations was around the first week of June 2023. Owing to this large delay, we focus on the resolution date instead.4

We observe excess returns reaching around 1.9% over the two weeks following the bipartisan Fiscal Responsibility Act (2023) passage by both houses of the government on 1 June 2023, as shown in Figure 4.

Figure 4. Positive perception after most recent debt deal

Problems loading this infographic? - Please click here

Source: Man Numeric.

The excess return over the S&P 500 index is close to the observed historical average of 1.5% from Figure 3 and further confirms our hypothesis that a stabilised outlook on government spending reflects positively for this basket.

What can we learn from this analysis? In summary, we conclude that being able to measure the sensitivity of stocks to government exposures could be informative not only for risk management, but also as a means of harvesting alpha.

 

With contributions from Diana Zheng, Quantitative Researcher, and Aditya Divekar, Associate Quantitative Researcher, at Man Numeric.

 

1. www.washingtonpost.com/business/2023/10/27/government-shutdown-mike-johnson/
2. Includes: 7 February 2014, 15 March 2015, 15 March 2017, 8 December 2017 and 1 August 2021.
3. Includes: 14 February 2013, 12 February 2014, 28 October 2015, 8 September 2017, 9 February 2018, 2 August 2019, 16 December 2021 and 1 June 2023.
4. 1 June 2023.

 

 

 

User Country: United States (237)
User Language: en-us
User Role: Public (Guest) (1)
User Access Groups:
Node Access Groups: 1