Why Yesterday's Market Rules Don't Always Apply Today

• Investors rely on market heuristics.
• Those heuristics are products of specific macro regimes, not universal laws.

• Successful investing depends on recognizing when the regime has changed.

   

Institutional investors routinely rely on cross-asset relationships to build portfolios, assess risk, and explain positioning. Many of those relationships become embedded in investment processes as simple heuristics:

• The 2-year Treasury yield tracks the federal funds rate.
• Rising front-end yields strengthen the dollar.
• Inflation lifts gold.

These rules of thumb work often enough to feel structural. They are not.

Rolling correlations across two decades of data show that each relationship strengthens, weakens, and sometimes reverses as macroeconomic conditions change. These breakdowns are not statistical noise around a stable long-run truth. They signal that the market is pricing a different source of uncertainty.

Cross-asset relationships are not fixed parameters. They are regime-dependent expressions of changing macroeconomic drivers. When the underlying regime shifts, heuristics often survive long after the mechanism that made them useful has disappeared.

For institutional investors, the challenge is not deciding whether a heuristic is right or wrong. It is recognizing when the conditions that made it reliable no longer exist. The three examples that follow illustrate why understanding those regime shifts is more valuable than relying on the heuristic itself.

Of the three relationships, this one should be the most stable. The 2-year Treasury yield is the market's running estimate of near-term policy, and the Fed sets the policy rate; the connection is almost mechanical.

Figure 1 shows that the correlation remains strongly positive across most periods but repeatedly turns negative around policy turning points. At those moments, the 2-year shifts from reflecting current Fed policy to pricing the next phase of the cycle—future cuts, slower growth, or an approaching inflection point. The breakdown is therefore diagnostic rather than anomalous. It signals that markets have begun looking beyond the current policy stance.

The logic is familiar: higher US front-end yields attract capital, lifting the dollar and US Dollar Index (DXY). The relationship is unstable across the sample and fails for three structurally distinct reasons.

The first failure is that DXY responds to rate differentials, not absolute yield levels. If US front-end yields rise but, say, European rates rise at a comparable pace, the marginal dollar advantage is unchanged.

The second failure is structural. Because DXY is heavily weighted toward the euro, periods of euro weakness can drive the index independently of movements in US front-end yields. Treating DXY as a pure dollar signal can therefore be misleading.

Risk sentiment provides a third channel. During periods of market stress, safe-haven demand can strengthen the dollar even as front-end yields fall. The relationship strengthens only when rate differentials, relative growth expectations, and risk sentiment reinforce one another. When those drivers diverge, the correlation weakens or reverses.

The third heuristic may be the most deeply embedded: Gold is an inflation hedge. When CPI rises, buy gold. Perhaps the strongest challenge to this view came from Claude Erb and Campbell Harvey in The Golden Dilemma, which found a correlation of approximately -0.82 between US real yields and gold prices. Their conclusion was that gold responds to the opportunity cost of holding a non-yielding asset, governed by real yields rather than headline CPI. The Federal Reserve Bank of Chicago's 2021 analysis using Treasury Inflation-Protected Securities (TIPS) yields confirmed the same finding.

Figure 3 confirms this empirically. From 2002 to 2013, the CPI-gold correlation was mostly positive, but the mechanism was not CPI itself. Nominal yields stayed contained while inflation moved, so real yields fell, and the cost of holding gold fell with them. The correlation turned negative from 2013 to 2017 because rising CPI now signaled normalization, higher real yields, and a stronger dollar. The heuristic produced exactly the wrong sign.

From 2018 onward, the relationship becomes unstable as gold is increasingly driven by real yields, safe-haven demand, central-bank purchases, and reserve diversification rather than CPI alone. Inflation still matters, but through its effect on real yields and macro uncertainty rather than as a direct driver of gold prices.

The breakdowns in these relationships are not independent failures. They reflect a broader shift in what markets are trying to price.

The first two heuristics share a common element: the 2-year Treasury yield. Yet the 2-year does not carry the same information in every environment. It always reflects expectations for monetary policy, but the forces shaping those expectations change as macro regimes evolve.

When policy is the dominant uncertainty, the 2-year closely tracks the expected path of the federal funds rate. When inflation disrupts that path, it becomes a vehicle for repricing policy expectations. Once that repricing is absorbed, the 2-year increasingly reflects US rate advantage relative to other major economies, reinforcing the dollar.

Figure 4 illustrates these changing roles. The dominant relationship shifts over time—not because the 2-year has changed, but because the source of macro uncertainty has changed. The heuristic remains the same while the mechanism underneath it evolves.

Post-2020 provides a clear example. Inflation first forced a repricing of expected Fed policy. Once that adjustment was largely embedded in markets, the front end increasingly reflected the persistence of U.S. policy relative to other central banks, reinforcing the dollar through widening rate differentials.

This is the broader lesson. Cross-asset heuristics are useful only when the mechanism that originally linked the variables remains intact. When the underlying driver changes, the heuristic may continue to look familiar while producing a very different outcome. The task for investors is therefore not simply to recognize a market relationship, but to identify the regime that gives that relationship meaning.

The practical implication is straightforward: investors should evaluate market relationships through the lens of the current regime rather than historical averages.

Three principles follow:

• Condition correlation analysis on regime. Historical averages combine periods in which relationships were driven by different economic forces and may misrepresent the current environment.
• Understand the mechanism before applying the heuristic. Ask what the market is pricing today before assuming a familiar relationship still holds.
• Reassess portfolio roles as regimes evolve. Assets such as gold may shift from inflation hedges to real-yield, reserve-diversification, or geopolitical hedges as macro conditions change.

Market relationships rarely disappear—they evolve. The challenge for institutional investors is not to memorize yesterday's heuristics but to recognize when the economic regime that gave rise to them has changed.