A major red flag for catastrophic losses is "thesis creep": repeatedly changing your reason for owning a stock as it declines. An investment made because it's a 'good business' at $10 becomes a 'value play' at $8, then a 'liquidation play' at $3. This intellectual dishonesty prevents cutting losses when the original thesis is broken.

Regularly re-evaluate your investment theses. Stubbornly holding onto an initial belief despite new, contradictory information can lead to significant losses. This framework encourages adaptation by forcing you to re-earn your conviction at regular intervals, preventing belief calcification.

A common mistake is altering an investment thesis to justify holding a losing stock, known as "thesis creep." A better approach is to sell immediately when the original thesis is proven wrong, rather than creating a new narrative to accommodate a falling price and avoid admitting a mistake.

Gaonkar identifies her biggest error with NVIDIA wasn't selling too early, but failing to re-evaluate and buy back in later. The psychological pain of "sunk cost bias" makes it incredibly difficult to re-enter a position at a higher price, even when the fundamental thesis has improved.

A successful investment can evolve into a different risk profile as it appreciates. For example, a cheap optionality bet can become a concentrated legal bet. Managers must recognize when a position has morphed out of their "wheelhouse" and have the discipline to exit, as the new risk factor may be one they are not equipped to manage.

To avoid confirmation bias and make disciplined capital allocation decisions, investors should treat every follow-on opportunity in a portfolio company as if it were a brand-new deal. This involves a full 're-underwriting' process, assessing the current state and future potential without prejudice from past involvement.

Post-mortems of bad investments reveal the cause is never a calculation error but always a psychological bias or emotional trap. Sequoia catalogs ~40 of these, including failing to separate the emotional 'thrill of the chase' from the clinical, objective assessment required for sound decision-making.

While chasing losses is a well-known trading pitfall, a more subtle danger is over-trading during a winning streak. The instinct is to double down and take more risk when flush with profits, but this is precisely when a trader should reduce exposure.

Gaonkar admits a major mistake wasn't just selling NVIDIA too early, but failing to re-evaluate it later. The sunk cost bias makes it psychologically difficult to revisit past decisions, especially ones that were wrong, causing investors to miss out on significant future gains.