The market is focused on inflation, but a deteriorating job market combined with high real rates could trigger a disinflationary spiral. Because the Fed is scarred by recent inflation, its response will be too slow, increasing the disproportionate chance that rates on the front end will have to return to zero to combat the downturn.
Gurevich opposes the mechanical application of stop-losses to every position. Risk management should be at the portfolio level. Some positions become more valuable as they move against you and should be held longer. A trader must preserve the freedom to exit a trade based on a changed thesis, not an arbitrary price level.
While seductive, complex trades with multiple conditions (knock-ins, knock-outs) create numerous ways for a core thesis to be correct on direction but still result in a loss. Simplicity in trade expression is a form of risk management that minimizes the pain of a good call being ruined by flawed execution.
To analyze a trade, ask what else must happen for your thesis to fail. For long Japanese bonds to lose money, the Bank of Japan must hike rates to 4%. A concurrent necessity of such a hike would likely be a massive rally in the Japanese Yen, which itself could be a hedge or a separate trade idea.
Beyond simple productivity gains, AI will eliminate the need for entire service-based transactions, such as paying for basic legal documents or second medical opinions. This substitution of paid services with free AI output can act as a direct deflationary headwind, a counterintuitive effect to the typical AI-fueled growth narrative.
Assets that grind higher over decades rarely reverse course suddenly. Instead, the prolonged, slow uptrend builds investor complacency, leading to an overbought state and a final parabolic surge. This blow-off top is the necessary precondition for a significant, sustained decline, a pattern Gurevich notes he missed in the bond market.
To compete with superior AI, a human trader should focus on building a portfolio of undervalued assets today. When hyper-intelligent AIs eventually arrive and re-price markets efficiently, they will likely buy the very assets the human trader already holds, validating the initial thesis and accelerating gains.
Select trades that are favorable under current market conditions but will also benefit from long-term secular trends if the initial thesis is wrong. This creates a resilient portfolio where if one part doesn't perform now, it's likely to become a valuable holding for a future market cycle, providing an embedded optionality.
Options are an excellent tool for risk management, not just speculation. When you have a high-conviction view that feels almost certain (e.g., "there is no way they'll hike"), buying options instead of taking a large vanilla position can protect the portfolio from a complete wipeout if your seemingly infallible view is wrong.