I am not sure if Kevin Warsh is a dedicated fan of Depeche Mode, but at the very least, he seems to agree with the sentiment of the band's 1990 classic song "Enjoy the Silence": sometimes, saying less is better. Warsh clearly applies this philosophy to what he considers the optimal communication strategy. This has left the market uncertain about what the Federal Reserve will do on Wednesday. The market currently prices in about a 9 basis point rate hike, but some are even preparing to bet on a 50 basis point increase.
Among the hawkish camp, there is a view that such a move would help strengthen the Fed's credibility. However, conversely, it could also be argued that a rate hike might damage that credibility. If a Warsh-led Fed were to actually raise rates, simultaneously delivering a fatal blow to the AI stock bubble, it would be both an interesting and deeply ironic situation.
Reports on the Fed's "internal debate" are quite intriguing. As the meeting minutes suggested, Warsh proposed forming several working groups without prior notice, which some committee members were clearly unhappy about. I don't know if this will make anyone more inclined to vote for a rate hike, but it at least indicates that some tension exists within the Federal Open Market Committee. Perhaps this is a good thing, as years of singing the tune of unity and cooperation have not brought inflation back to target.
The lack of policy guidance and informal leaks have led some market participants to speculate that the Fed will raise rates this month to curb renewed inflationary pressures and restore a degree of credibility. Given that we have very little empirical data to judge the Fed's actions, this rationale for a rate hike is not without merit. However, I believe it cannot be taken for granted that a rate hike will necessarily enhance the Fed's credibility.
What is "central bank credibility"? From at least one perspective, it means the central bank can act in a rational, consistent manner in response to economic changes. In other words, it means following a predetermined policy reaction function. This is the primary purpose of the quarterly Summary of Economic Projections and the dot plot. Warsh is correct that FOMC members are not good at forecasting, and it would indeed be ridiculous if the market took their predictions for the economy and interest rates at face value. Still, we can discern a logic from them: if X happens, the Fed will take Y.
In recent years, this column has modeled and analyzed the Fed's policy reaction function after each SEP release. The last projection showed the Fed's sensitivity to inflation had increased significantly compared to March. Given the changes in CPI and PCE data since the outbreak of the Iran conflict, this is understandable. However, the problem is that since the last Fed meeting, the year-on-year PCE price index has been in line with expectations, and both CPI and PPI have been very moderate. The Fed's SEP forecast of 3.6% inflation by year-end now seems too high. Admittedly, oil prices are still higher than a few weeks ago, but they have not yet caused a significant inflationary impact. In fact, the national average gasoline price in July was still slightly below the June average.
You could also argue that central bank credibility depends not only on adhering to a set reaction function but also on behavioral factors. If the president constantly clamors for rate cuts, the best way to demonstrate central bank independence might be to raise rates. However, several FOMC members, including Christopher Waller who is skeptical of the working groups, have proposed their own judgment frameworks: if inflation trends do not improve, they would support a rate hike. But the June data suggests inflation may indeed be cooling. If they vote to raise rates on Wednesday, it would conflict with the reaction function they previously set for themselves. But sometimes, people change their minds. One of Warsh's legitimate criticisms of forward guidance is that it can lead people to become overly anchored to forecasts that may be outdated, or even to an outdated policy reaction function. Even so, if someone votes to raise rates on Wednesday, they should at least explain why their reaction function has changed since the last meeting.
These factors are particularly noteworthy now, as the AI bubble shows signs of cooling. The Korea Composite Stock Price Index tumbled again on Tuesday, erasing more than 61.8% of its gains since 2026. Based on the closing price, the index is down nearly 34% from its peak, the largest drop since the COVID-19 pandemic. In the context of the past 45 years, this is clearly a significant downtrend. Of course, the Korean stock market has no direct relationship with Fed policy, and it shouldn't. But it is closely linked to an important area of the US financial market and economy: the AI investment boom.
To be fair, the US market has not yet experienced the same pressure as the Korean market. In fact, the ADR of SK Hynix still maintains a considerable premium over the dollar value of its underlying stock. The US economy and stock market returns depend on the growth of the AI industry. An obvious risk of a "credibility-restoring" rate hike is that it could exacerbate financial stress on this engine. Warsh might well say, "So what?" He likely believes the Fed has no responsibility to underwrite stock market gains or to intervene when the market overextends itself. Borrowing from another song on Depeche Mode's 1990 classic album *Violator*, Warsh might call this approach "Policy of Truth." At the very least, it is reasonable to infer that if the FOMC's policy reaction function shifts significantly hawkish, the correlation between US stocks could rise, amplifying volatility at the S&P 500 level.
I still doubt this scenario will occur. However, people are people, and they do not always act strictly rationally.
Comments