Retail individual investors are increasingly being targeted with complex investment products based on derivatives. Such investors, with their limited knowledge of finance, might not fully appreciate the variability of returns associated with so-called structured products. Sensitivity analysis is often recommended as a way for investors to explore the variability of returns and thus make better informed investment decisions. Through sensitivity analysis, investors are expected to assess investment products by examining them under different scenarios depicting the uncertain conditions that affect the returns from these products. However, a sensitivity analysis is only as valuable as the ranges over which such analysis is conducted. The theory of framing in behavioral decision theory suggests that the range employed for sensitivity analysis can significantly influence an investor’s assessment of a financial product.
In this paper, we report the results of an experiment where two groups of subjects were exposed to an investment product linked to movements in the inter-bank interest rate. Subjects were instructed to carry out sensitivity analyses with respect to this rate using Microsoft Excel. The two groups faced identical experimental stimuli except for a single difference: they conducted their sensitivity analyses over different ranges of the inter- bank interest rate. As the theory of framing predicts, the same investment product was viewed differently by the two groups. Our results suggest that the range over which sensitivity analysis is conducted can be manipulated to influence how an investment product is perceived, thus undermining the efficacy of such analysis. When this happens, private investors unfamiliar with investment products run the risk of making inappropriate investment decisions with the potential for significant personal and social costs.
Das, Amit; Chua, Cecil Eng Huang; and Seal, Kala, "THE SENSITIVITY OF SENSITIVITY: HAZARDS OF SOFTWARE-BASED FINANCIAL DECISION MAKING" (2007). ICDSS 2007 Proceedings. 10.