Climate risk brings about a new type of financial risk that standard approaches to risk management are not adequate to handle. We develop a model that allows to compute the valuation adjustment of corporate and sovereign bonds conditioned to climate transition risk, based on available forward-looking knowledge on climate policy scenarios provided by climate economic models. We investigate the impact of the endogeneity and deep uncertainty of future scenarios on both the valuation of individual bonds and on standard risk metrics for a leveraged investor, considering the role of fossil fuels and carbon intensive activities in the economy of countries. We demonstrate that an investor’s Value at Risk andExpected Shortfall have low sensitivity to key parameters such as the Probability of Default(PD) of the bond In contrast, the investor’s PD is very sensitive to these parameters, and increases with the PD of the bond and with the probability of occurrence of the adverse climate transition scenario. Choosing the wrong scenario could lead to a massive underestimation of losses. Thus, Climate stress test exercises should allow for a wide enough sets of scenarios to avoid underestimation of losses. We apply the methodology to the AustrianNational Bank’s portfolio of sovereign bonds. In carbon intensive countries, the cost of climate misalignment is reflected in a higher Climate Spread and affects sovereign risk and portfolio’s performance. These results have important implications for the selection of relevant climate transition scenarios in climate stress-testing exercises, and for the assessment of climate-related financial risk for supervisory and prudential policy purposes.