Graduation Year

2022

Date of Submission

12-2021

Document Type

Campus Only Senior Thesis

Degree Name

Bachelor of Arts

Department

Economics

Reader 1

Darren Filson

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Abstract

Two phenomena are threatening the relevancy/importance of Silicon Valley (SV), the tech capital of the world, and home of many of the world’s most valuable companies. First, the SV Tech Exodus is causing major tech firms to decrease, or fully eliminate, their real estate presence in SV. Second, Remote Work is causing tech firms to abandon their traditional work models and opt for fully-remote or hybrid workforces. This study uses a stock-market event study to estimate the equity-value implications of a firm announcing Exodus or Remote Work decisions.

Using the Fama-French Five Factor Model, I estimate a model of normal returns to calculate Cumulative Abnormal Returns (CARs) in the days surrounding an announcement. I examine three types of announcements for the Tech Exodus: Headquarters (HQ) Relocation out of SV, SV Downsizing, and adding a Second HQ outside SV. For Remote Work, there are two categories of announcements: supporting fully-remote work, and supporting a hybrid model. I hypothesize that firms who announce a HQ Relocation will experience positive CARs (H1), firms that announce SV Downsizing will experience positive CARs (H2), and that HQ Relocation announcements will have larger positive CARs than SV Downsizing announcements (H3). I hypothesize that firms that announce a decision to support fully-remote work will experience positive CARs (H4), and that for comparable announcements of remote work, pure software companies will experience more positive CARs than other tech companies (H5). Finally, I hypothesize that Exodus announcements will have a larger absolute effect on CARs than Remote Work announcements (H6).

I find support for H1: firms that announce a HQ Relocation experience a mean CAR of 2.95% or 3.76% (significant at 10% and 5%, respectively) across a two-day event window. The two different CAR values pertain to a model that considers Tesla and a model that excludes it. I reject H2: firms that announce SV Downsizing experience a mean CAR of -3.59% across a three-day event window (significant at the 5% level). Given the difference in sign between HQ Relocation and SV Downsizing, the initial conditions for H3 do not hold, but the result is consistent with the hypothesis. I find evidence that firms that announce a decision to support fully-remote work experience a mean CAR of -0.78% and 0.01% across a three-day and two-day event window, respectively. These findings are not statistically significant, so they do not provide strong evidence for or against H4, but the point estimates suggest I reject the hypothesis. I reject H5. The mean absolute value of Exodus CARs is greater than that of Remote Work, providing evidence for H6, but the difference between those two values is not statistically significant.

Taken at face value, the findings of the Exodus study suggest that if a firm is going to be in SV, it should be all in and maintain its presence. If a firm is trying to shrink its SV presence, it should relocate out entirely instead of downsizing. The insignificance in results for the Remote Work question may mean that the benefits of remote work are about equal to the costs. Of course, self-selection cannot be overlooked: it remains possible that these findings cannot be useful for providing recommendations to other firms. Although there are several limitations, and many opportunities for next steps, this study begins to help us understand the future of Silicon Valley and Remote Work.

Download

This thesis is restricted to the Claremont Colleges current faculty, students, and staff.

Share

COinS