Commercial forestland in the U.S. is unique enough in terms of risk/return profile and large enough in terms of aggregate value to warrant consideration for inclusion as a distinct asset class in the pension asset allocation process. Returns to investments in timber can be calculated from publicly available prices for end products, estimates of monthly cash flows (derived from a model that takes into account initial investment, operating costs, volume and biological growth), estimates of the value of standing timber and of the value of the underlying land.

Returns will depend on the harvest strategy employed. Return series were generated under the assumptions that a timber portfolio consisted of N acres, with each acre having timber of different ages and timber being cut at age N. Thus a 25-year harvest strategy would have 25 acres of timber ranging in age from zero to 24 at the beginning of the year and one to 25 at the end (assuming timber is harvested at the end of the year and replanted at the beginning of the following year).

Monthly returns and standard deviations of return for harvest strategies of 10 to 41 years show that strategies of 23 years and up dominate. Most importantly, the data expose a 10 to 11-year window of potential optimal harvest. During this window, investment grows in volume and quality. A second return series reflects the results of selling during this window only when prices are 5 per cent above the rolling 12-month trend line for timber prices. This produces higher returns and lower variances than the first return series.