Ai As A Macro/Market-Structure Disruptor And An Epistemic Constraint
Key takeaways
- Alex Gurevich argues that AI can be locally anti-growth and deflationary by eliminating entire paid activities (e.g., legal drafting and second medical opinions) rather than reallocating them into new spending.
- Alex Gurevich ranks trades using carry as only one input alongside trend, valuation, and links to global growth/technology.
- Alex Gurevich says he made a major mistake by not recognizing post-2020 breakdown risk signaled when long-bond futures broke upward out of a decades-long channel during the COVID rally.
- The second edition of The Next Perfect Trade preserves the original 2015 text and adds 2025 commentary to evaluate how the principles performed and to reduce selective success bias.
- Alex Gurevich says that in early 2022 he bought options on Eurodollar/interest-rate contracts to limit downside versus a potentially catastrophic rates selloff.
Sections
Ai As A Macro/Market-Structure Disruptor And An Epistemic Constraint
AI appears in three roles: (1) a contested conviction claim that AI skepticism is misguided, (2) an expectation that AI will displace unaugmented discretionary trading over time (including longer horizons), and (3) a proposed macro channel where AI can be locally deflationary by eliminating paid services rather than reallocating into new spending. Separately, AI is presented as an explicit limit on long-horizon forecasting reliability under an AI-transcendence scenario, motivating process adaptation (building internal AI platforms) rather than precise long-dated predictions.
- Alex Gurevich argues that AI can be locally anti-growth and deflationary by eliminating entire paid activities (e.g., legal drafting and second medical opinions) rather than reallocating them into new spending.
- Alex Gurevich says AlphaGo and subsequent 'zero' engines convinced him by 2016 that AI can discover creative patterns humans cannot, beyond brute-force calculation.
- Alex Gurevich asserts that skepticism about the power of AI is misguided and says he began laughing at such skepticism nearly a decade ago.
- Alex Gurevich argues that AI transcendence is inherently hard for an unaugmented human to predict, limiting reliable long-horizon macro forecasting under an AI-singularity scenario.
- Alex Gurevich expects AI will soon displace unaugmented discretionary trading.
- Alex Gurevich says he is building internal AI platforms to augment his discretionary process and expects AI capital flows to eventually crowd into assets he already views as undervalued.
Liquidity-Regime Mapping And Cross-Asset Positioning Logic
Gurevich presents a simplified liquidity regime map (tight vs high liquidity) to guide broad exposure choices, while also emphasizing that carry is subordinate to other factors like trend and valuation. Multiple examples illustrate that carry can help (oil backwardation) but does not guarantee success (JPY/CHF), and that negative-carry assets (precious metals) can still be central if the broader setup warrants it.
- Alex Gurevich ranks trades using carry as only one input alongside trend, valuation, and links to global growth/technology.
- Alex Gurevich says that in recent years oil offered positive carry via backwardation where deferred contracts were cheaper than spot.
- Alex Gurevich gives an example that carry-positive FX setups can fail, saying yen versus Swiss franc did not work well in 2025 despite favorable carry.
- Alex Gurevich asserts that avoiding negative carry would exclude holding precious metals and would have missed much of gold’s move, and he also asserts that spot gold led mining stocks until 2025.
- Alex Gurevich says he exited an oil carry trade in 2025 as oil began grinding down and the setup stopped feeling compelling despite earlier carry gains.
- Alex Gurevich frames trade selection around a binary liquidity regime: if liquidity tightens he prefers long bonds and long USD versus low-yield currencies, while if liquidity is high he prefers precious metals and other liquidity-sensitive assets.
Macro Regime Breaks: Bonds, Stock-Bond Correlation, And Zirp Reversion Risk
The corpus highlights two regime-related observations: (1) a perceived technical regime signal in long bonds around the COVID rally and (2) the 2022 breakdown of the typical equity/bond diversification relationship. On the forward-looking side, it contains a coherent but unresolved thesis that labor-market deterioration plus positive real rates could lead to disinflation and potentially front-end rates returning toward zero, with an added mechanism that inflation fixation can delay policy response due to data lags.
- Alex Gurevich says he made a major mistake by not recognizing post-2020 breakdown risk signaled when long-bond futures broke upward out of a decades-long channel during the COVID rally.
- Alex Gurevich cites 2022 as an episode where both equities and bonds sold off together, breaking the classic risk-parity assumption that one rallies when the other sells off.
- Alex Gurevich argues that if the equity-market-driven financial impulse fades while job markets decay, interest rates could proportionally return toward zero.
- Alex Gurevich argues that consensus fixation on inflation makes disinflation more likely because inflation fear delays policy response and inflation data are lagging, increasing odds of the Fed and fiscal policy being behind the curve.
- Alex Gurevich uses the heuristic that very long, slow grinding uptrends typically end with a parabolic rally before a meaningful downtrend can begin.
- Alex Gurevich expects a new bull market in U.S. Treasuries and sees a disproportionate chance that front-end U.S. rates eventually return toward zero because positive real rates plus a deteriorating jobs market can trigger a disinflationary spiral.
Trade-Construction Robustness And Scenario Invariants
A recurring theme is designing trades around robustness: positions that can work across time horizons and scenarios, using scenario 'necessities' and the alignment of contextual 'location' with narrative 'story'. The second-edition structure (original text plus retrospective commentary) is presented as a method to audit principles over time and avoid selective success bias.
- The second edition of The Next Perfect Trade preserves the original 2015 text and adds 2025 commentary to evaluate how the principles performed and to reduce selective success bias.
- Alex Gurevich wrote The Next Perfect Trade to define ex-ante characteristics of a good trade independent of whether one uses fundamental or technical market-direction methods.
- Alex Gurevich uses a 'sort of necessity' framework that asks what must also be true if the market’s priced scenario occurs, and he prefers trades that work across a broader range of scenarios.
- Alex Gurevich evaluates trades using 'location' (valuation/range context) and 'story' (narrative tailwinds), with the best trades occurring when both align.
- Alex Gurevich prefers trades that can plausibly make money either immediately or later, rather than requiring a single precise timing outcome.
Risk Management And Instrument Choice Under Uncertainty
The corpus emphasizes pragmatic execution: trend participation without full causal explanation, skepticism about complex derivatives unless tightly justified, and explicit acknowledgment of options path-dependence. Options are framed as a tool for capping tail risk when outcome distributions are uncertain (example given in rates in early 2022), while stop losses are disfavored in favor of discretionary exits tied to conviction.
- Alex Gurevich says that in early 2022 he bought options on Eurodollar/interest-rate contracts to limit downside versus a potentially catastrophic rates selloff.
- Alex Gurevich sometimes chooses to stop trying to fully explain price moves and instead participates in trends while applying risk management.
- Alex Gurevich argues that options and complex derivatives add failure modes where one can be directionally right yet still lose money, so options require a strong and specific rationale.
- Alex Gurevich highlights a common options failure mode: the underlying moves in the correct direction but the move is insufficient versus strike and time horizon, leading to losses.
- Alex Gurevich says he dislikes stop losses because he prefers discretion to exit trades when conviction deteriorates rather than being forced out by a price level.
Watchlist
- Alex Gurevich suggests that in a 'sell America' environment where everything in the U.S. falls together, investors may need to reframe necessities and potentially look to other markets such as Japan.
Unknowns
- What concrete indicators or thresholds does Gurevich use to declare that equities have become 'much cheaper' (and over what timeframe)?
- What specific labor-market deterioration signals (levels, momentum, breadth) would, in his framework, trigger the 'rates toward zero' pathway versus a different policy outcome?
- How does Gurevich operationalize 'liquidity' (which measures, which frequency) when applying his binary liquidity-regime map?
- What are the concrete assets or factors Gurevich considers undervalued that he expects future AI-related flows to crowd into?
- What evidence would demonstrate that AI is displacing discretionary trading at the time horizons he discusses (e.g., measurable changes in market microstructure, strategy performance dispersion, or flow composition)?