This thesis proposes to verify if there is evidence of an inflation-betting effect in the Brazilian government bonds market and if it interacts with the aggregate supply of government bonds. We propose an empirical study about the relationship between inflation disagreement, measured as the standard deviation of one year ahead inflation forecasts, and excess holding returns of government bonds in different supply scenarios - based on the works of Hong et al. (2016) and Greenwood and Vayanos (2014). We analyzed two types of Brazilian government bonds, zero-coupon LTNs and NTN-Bs, from January 2005 to December 2017. Our sample contains 1,809 monthly observations of annual excess holding returns, government bonds’ supply, maturity and stock held by the Brazilian Central Bank. The results suggest that in Brazilian market the expected effect based on liquidity premium theory does not apply. If exposed to higher levels of uncertainty about future inflation, investors should demand a risk premium in order to invest in long-term bonds, but the opposite seems to happen: the coefficient of inflation disagreement was negative, decreasing with maturity and presented a more pronounced effect for NTN-Bs, inflation-linked securities that have longer maturities. Moreover, when analyzing scenarios of low and high aggregate supply, we find that the negative effect of inflation disagreement is more intense in scenarios of low supply, both for LTNs and NTN-Bs. These findings point to the existence of an inflation-betting effect in Brazilian government bonds market. In times of heterogeneous expectations, investors speculate on future inflation and prefer to bet on long-term bonds, that are more sensitive to inflation than short-term ones. This effect becomes more pronounced in times of low aggregate supply, when short-term constraints are binding.