This article uses four types of financing rates to test for market efficiency and finds that each of the four carrying costs is not sufficient to eliminate arbitrage opportunities for the period examined. It is found, however, that the inflation hypothesis is able to explain a significant part of observed market inefficiencies. By regressing the futures-forward rate differential on the inflation rate, the inflation hypothesis postulated to explain the futures-forward rate disparity is that (i) the longer the arbitrage period, the more likely the parameter estimate is to be significant; and (ii) the more the inflation-certain financing instrument is used, the lower the significance of the parameter estimate regardless of the arbitrage period. Regression analyses by grouping time to futures' expiration date support the inflation hypothesis of (i) and (ii) for financing rates using shorter-term T-bill, 90-day T-bill, and federal funds rates, while the result is mixed for the case using overnight repo rates by which the impact of inflation alternates over the groups. Shorter-term T-bill, 90-day T-bill, and federal funds rates are found to follow a first-order homogeneous nonstationary process (i.e., the first-order difference is stationary), but the overnight repo rate is not. The relatively high nonstationary of the overnight repo rate is suspected to be a contributing factor to its mixed result.