A risk manager is performing a backtest on a company's 1-day 99.5% Value at Risk (VaR) model over a 10-year period, utilizing a 95% confidence level. The firm operates on 250 trading days each year, and the daily returns are assumed to follow an independent and identical distribution. Considering these parameters, what is the closest maximum number of daily losses that exceed the 1-day 99.5% VaR within a 10-year span, which would suggest that the model is accurately calibrated?