Amundsen versus Scott: are growth paths related to firm performance?

Abstract

In the race to the South Pole, Roald Amundsen’s expedition covered an equal distance each day, irrespective of weather conditions, while Scott’s pace was erratic. Amundsen won the race and returned without loss of life, while Scott and his men died. In the context of firm growth, the Amundsen hypothesis suggests that smoother growth paths are associated with better performance in subsequent periods. We develop a new method to investigate how firms’ sales growth deviates from their long-run average growth path. Our baseline results suggest that growth path volatility is associated with higher growth of sales and profits, but also with higher exit rates. However, this result is driven by firms with negative growth rates. For positive-growth firms, volatility is negatively associated with both sales growth and survival, providing nuanced support for the Amundsen hypothesis.

Plain English Summary

In the race to the South Pole, Roald Amundsen and Robert Falcon Scott adopted different strategies that resulted in victory for Amundsen and death for Scott. Amundsen’s approach was to consistently pace his team (to cover a fixed and equal distance each day), while Scott sought to cover as much distance as possible each day. In the context of firm growth, this relates to the tradeoff between steady growth (low volatility in growth rates) and growing as fast as possible in each period (potentially leading to high volatility in growth rates). We develop a new set of indicators for quantifying firms’ growth paths, observing that growth path volatility in general is associated with higher growth of sales and profits, but also with higher death rates. This result is driven by firms with negative sales growth, however. Like Amundsen, it seems beneficial for firms with positive sales growth to pace themselves to increase their subsequent growth and likelihood of survival.

Appendix 1 details on the sample and variables

We also test if our results are robust to commonly used growth dispersion measures, such as the absolute min–max growth difference (AMM). The AMM measure is calculated by first ranking a firm’s growth rates in order of size and then calculating the absolute difference between the firm’s largest growth rate and the smallest growth rate. More formally, if $\left|{s}^{\mathrm{max}}\right|=\mathrm{max}{\left(\left|\mathrm{ln}\frac{S\left(t+1\right)}{S\left(t\right)}\right|\right)}_{t=0}^{T}$ and $\left|{s}^{\mathrm{min}}\right|=\mathrm{min}{\left(\left|\mathrm{ln}\frac{S\left(t+1\right)}{S\left(t\right)}\right|\right)}_{t=0}^{T}$ corresponds to the largest and smallest absolute growth rates over the sequence of growth rates, the AMM measure can be defined as:

$$AMM=\left|{s}^{\mathrm{max}}-{s}^{\mathrm{min}}\right|$$

Finally, in order to compare our results with a traditional volatility measure, we have also done all estimations using the standard deviation (SD) as our growth dispersion measure.

All growth dispersion measures that are used in the paper are normalized to have mean zero, and a standard deviation of one. To calculate $Area$ and $AMM$, we use consecutive periods, each with $T=4$. The outcome measures, we consider is the future growth rate of $\mathrm{ln}\frac{S\left(T+k\right)}{S\left(T\right)}$ with $k=\mathrm{1,2}$ and $3$.

The correlation between our new indicator of growth dispersion and the more traditional ones are presented in Table 1

Our main results remain qualitatively similar when using the alternative growth dispersion measures. Note, however, that these alternative growth dispersion measures cannot be used to categorize firms into different growth types (see Table 5).

Table 5 Correlation matrix

Table 6 Definitions of variables