Breeding new crop varieties with resistance to the biotic stresses that undermine crop yields is tantamount to increasing the amount and quality of biological capital in agriculture. However, the success of genes that confer resistance to pests induces a co-evolutionary response that depreciates the biological capital embodied in the crop, as pests evolve the capacity to overcome the crop's new defences. Thus, simply maintaining this biological capital, and the beneficial production and economic outcomes it bestows, requires continual reinvestment in new crop defences. Here we use observed and modelled data on stripe rust occurrence to gauge changes in the geographic spread of the disease over recent decades. We document a significant increase in the spread of stripe rust since 1960, with 88% of the world's wheat production now susceptible to infection. Using a probabilistic Monte Carlo simulation model we estimate that 5.47 million tonnes of wheat are lost to the pathogen each year, equivalent to a loss of US$979 million per year. Comparing the cost of developing stripe-rust-resistant varieties of wheat with the cost of stripe-rust-induced yield losses, we estimate that a sustained annual research investment of at least US$32 million into stripe rust resistance is economically justified.