|

Stress Test Result

Verdict:

ROBUST

How your thesis holds up against the knowledge base

In a regime where inflation and nominal yields are high enough that inflation (not growth) is the dominant macro shock, stocks and bonds can sell off together; therefore a classic 60/40 becomes structurally less reliable and should be rebalanced (e.g., less duration reliance, add an inflation/real-asset sleeve).

Yes: when inflation is the constraint, your ‘diversifier’ stops diversifying and starts helping equities dig the hole. Just don’t confuse a regime framework with a permanent obituary for bonds—correlation is conditional, not a marriage vow.

Thesis

Rebalance away from 60/40 — stock-bond correlation flips positive in high inflation regimes

Source: Popular macro take

Attack Vectors

Where the knowledge base challenges your reasoning

You’re trading a regime claim without proving the regime line is actually crossed (or still crossed).

minor

Your mechanism only works when inflation/yields are “high enough to matter” such that the marginal shock is inflationary. If inflation is already re-anchoring (or the market’s dominant fear flips back to recession), bonds can regain their crash-hedge behavior and your ‘rebalance away’ becomes a late-cycle whipsaw. The whole point is regime-dependence—so if you can’t specify what ‘high enough’ means in observable terms, you’re just narrating 2022 forever.

📚 1970s: inflation shocks drove simultaneous pain in stocks and bonds; correlation was not your friend

📚 1994: bond selloff hurt diversified portfolios even without an equity crash—rates were the shock

📚 2022: classic 60/40 drawdown as inflation forced tightening and both assets repriced lower

📚 2013 taper tantrum: rates shock hit duration and risk assets together, briefly flipping correlation dynamics

“Rebalance away from 60/40” is directionally right but operationally vague—what replaces the bond hedge when it stops hedging?

minor

The thesis correctly identifies the failure mode (positive correlation in inflation regimes) but can still lose money if implemented as a blunt ‘sell bonds’ rather than a redesign. The playbook in an inflation-dominant regime is typically a third pillar (commodities/energy/gold/TIPS/short duration/cash-like optionality) because the problem isn’t ‘bonds bad’—it’s ‘duration as hedge breaks.’ If you don’t specify the replacement ballast, you’ve just reduced diversification and called it risk management.

📚 1970s: US stocks and long bonds delivered poor real returns; commodities/energy outperformed

📚 1940s US: bondholders faced negative real returns under inflation and repression; real assets helped preserve purchasing power

📚 2000–2010: commodities and energy had strong runs while 60/40 had regime-specific challenges (especially early 2000s equity bear)

📚 2022: classic 60/40 drawdown as inflation shock hit both legs simultaneously

Correlation flips can be episodic around rates shocks; you can overfit a ‘structural’ story to what might be a transition phase.

minor

Even within the framework, correlation dynamics can flip “briefly” during rates shocks (not just in a decade-long inflation era). If you treat every rates-driven drawdown as proof that 60/40 is dead forever, you risk abandoning duration right as inflation cools and the hedge property returns. The memory version of this is simple: the diversifier can change sides—meaning it can also change back.

📚 1970s: persistent inflation produced positive stock-bond correlation for long stretches

📚 2000–2019: disinflationary regime restored negative correlation and made 60/40 look like genius

📚 2022: inflation shock drove one of the worst simultaneous stock/bond drawdowns in modern history

📚 1965–1981: long bond bear market coincided with equity valuation pressure from inflation/real-rate uncertainty

Evidence Gaps

What you should verify before putting money on this

📊No explicit verification that inflation and nominal yields are currently above the regime threshold where inflation shocks dominate growth shocks.

→ What to check: Is the market’s dominant fear inflation persistence (vs recession), and are inflation/yield levels still in the ‘inflation binds policy’ zone rather than the ‘policy can ease’ zone?

📊No timing plan for how long the correlation regime is expected to persist before reverting.

→ What to check: What would constitute “inflation convincingly re-anchored” (several quarters of benign inflation) such that bonds plausibly regain negative correlation—and are you monitoring for that?

Hidden Assumptions

What your thesis needs to be true to work

  • This thesis requires inflation and yields to be above the regime threshold where inflation shocks dominate growth shocks in asset pricing (otherwise bonds can revert to hedging equities).
  • This thesis requires you not to assume (as many model portfolios implicitly do) that Treasuries reliably provide negative correlation to equities across regimes—i.e., correlation is not a promise, it’s conditional.
  • This thesis requires inflation to remain the binding constraint on policy (limiting the central bank’s ability to ease into growth shocks), rather than quickly re-anchoring into a ‘policy can ease’ zone where duration resumes its hedge role.
What Would Change My Mind

Conditions that would upgrade or downgrade this verdict

Upgrade if:

  • Multiple quarters of inflation that remain persistently upside/two-sided, with repeated episodes where hot inflation/rates shocks coincide with simultaneous equity and duration drawdowns.
  • Clear evidence that policy is constrained by inflation (i.e., growth scares don’t reliably produce easing expectations/rallying duration), keeping stock-bond correlation unreliable in stress.

Downgrade if:

  • Several quarters of benign inflation that convincingly re-anchor the regime and restore policy room to ease, followed by equities selling off while Treasuries rally (negative correlation returns).
  • A sustained shift in the market’s dominant shock from inflation to growth/recession such that bad growth data consistently pulls yields down and cushions equities.

Updated your thesis with new evidence? Run it again.