Long volatility positions lose money most of the time. They pay theta, they suffer vix-roll-yield if expressed in futures, and they need either a genuine move or a repricing of fear to work. Their justification is convexity: the payoff in the rare good scenario is far larger than the accumulated cost.
This is why long volatility is usually run as an overlay rather than a standalone strategy. Sized as a few percent of a portfolio's annual budget, it converts a catastrophic drawdown into a manageable one; sized as the main position, it bleeds to nothing waiting.
Example: buying XYZ 45-day strangles every cycle costs about $1.20 a month. Ten quiet months cost $12. The eleventh month gaps 25% and the strangle is worth $11. The strategy roughly breaks even over the year while having been profitable in exactly one month.
Related: short-volatility-trade, left-tail-hedge, vega-convexity, backspread