When historical averages are needed, an Olympic average often is used rather than a simple average in calculating benchmarks in farm bill commodity programs. For example, the Agricultural Risk Coverage (ARC) program that was passed by the Senate Agriculture Committee uses Olympic averages of prices and yields in calculating benchmark revenue. In this post, Olympic averages are compared to simple averages for corn and soybean prices. Generally, Olympic and simple averages will track one over time. The relationship of Olympic to simple averages depends on the nature of distributions across time.
Olympic averages eliminate the high and low observations and then average all remaining observations. The simple average uses all observations when it takes its average. The nature of outliers will determine whether the Olympic average is above or below the simple average. When all observations are in a narrow range except for one outlier that is below the remaining observations, the Olympic average will be above the simple average. This occurs because the Olympic average will eliminate the low outlier. On the other hand, when observations are in a narrow range and one observation is above the remaining observations, the Olympic average will be below the simple average. More complicated differences in outliers will result in unpredictable relationships between the Olympic and simple average.