Modeling the Response to Exogenous Shocks: The Capital Uplift Rate in Petroleum Taxation

Firms are exposed to exogenous shocks, such as sudden shifts in exchange rates, input and output prices, corporate law, and the tax system. Regulators, financial analysts, and researchers who want to understand how such shocks affect the value of the firm must make a fundamental methodological choice: Can they just assume the firm’s behavior does not respond to the shock (unresponsive firm), or should they instead assume that behavior changes (responsive firm)? 

This choice involves a tradeoff between simplicity and realism. Assuming an unresponsive firm is the simpler and easier approach, because assuming a responsive firm requires a specific idea or model of how the firm will react. Such responsive models may be difficult to build and even harder to implement. On the other hand, responsive models may better predict what the firm will eventually do when the shock occurs.

In 2013, the level of Norwegian petroleum taxation was increased by reducing the annual capital uplift rate from 7.5% to 5.5% of capital investment per year over the first four years of the field’s life. Capital uplift is extra depreciation deducted from taxable income in order to protect normal returns (taxed at 28%) from being taxed as abnormal returns (taxed at 78%). The tax change we study is large. For instance, the tax shield from capital uplift used to be 30% (ie, 4 times 7.5%) of capital investment in all planned fields on the Norwegian continental shelf and about 35% of shareholder value in the fields we consider in this paper. 

Our paper analyzes how this tax shock has behavioral effects (eg, on the capital investment, extraction rate, and production period) and value effects (on the net present value of the shareholders’ residual claim after taxes) in a wide range of petroleum fields.

We find that both the behavioral effect and the value effect are quite large if we assume the taxpayer responds optimally. Surprisingly, however, the value effect is quite similar if we instead assume the taxpayer does not change behavior after the tax shock. 

Therefore, the improved insight gained by modeling a responsive rather than unresponsive taxpayer may not be worth the effort if the primary concern is to understand value effects rather than behavioral effects. Specifically, we find that, although optimal design changes considerably and field value drops by 12%, the ability to reoptimize design after the tax shock is only worth 1.5% of field value. 

This evidence suggests that large behavioral effects of a regulatory change do not necessarily imply large value effects, making it less important to always account for the taxpayers’ response. The valuation error in such cases may be moderate if one instead uses the simplifying and widespread assumption of unresponsive taxpayers. 

Our paper is available at here.

Magnus Berg is Equity Analyst at Arctic Securities, Øyvind Bøhren is Professor of Finance at BI Norwegian Business School and Erik Vassnes is Deputy Director General at Ministry of Finance, Norway.