There are four things we have been pointing to over the past few days that could still send this market circling the drain. That is hardly a pleasant thought, but it is reassuring, in a strange way, to know we are not the only ones looking past the peace deal confetti and reaching for a seat belt.Takeaways The peace deal removed the fat tail, but not the market’s structural fragility. War risk has shrunk, but equities now have to trade on their own plumbing again.Goldman’s Brian Garrett is flagging the key market tell: this rally still looks powered more by short covering than fresh long conviction. That can start a move, but it cannot sustain one forever.Kevin Warsh’s first Fed test matters because the peace dividend can quickly run into a rates wall. If the market hears hawkish credibility instead of dovish relief, the rally gets harder to extend.AI policy risk and record IPO supply are now the bigger market stress tests. If Washington clouds the AI leadership trade while SpaceX, Anthropic and OpenAI pull capital toward new paper, the market premium still has to earn its keep. Party PoopersJan-Patrick Barnert at Bloomberg Markets Live frames the issue cleanly: pricing out a Middle East war may be the easy part for equities. The harder part is what comes next, because the market still has to navigate three much bigger domestic and structural risks: a potentially hawkish Fed chair, Washington’s intervention in the AI trade, and the largest wave of equity supply in market history.The geopolitical tail has thinned, and that matters. The US and Iran are due to sign an interim peace agreement on June 19, with the possibility that the Strait of Hormuz fully reopens. Brent has already reversed roughly 80% of the war-driven jump, which tells you the market has largely cashed the peace cheque. But Barnert is right to caution that this is not the same as a lasting peace. The April truce wobbled within days, Hormuz traffic may only normalize gradually, and the agreement still needs ink. The war risk has shrunk, not disappeared. That is why the market is already shifting its attention back to the problems equities must solve without the help of geopolitical relief.The first problem is positioning. Brian Garrett, a derivatives specialist on Goldman Sachs’ trading desk, points out that the early-summer tone has been built more on macro short-covering than on genuine conviction. His key point is that crowded one delta hedges have been unwound into a market still searching for direction. That is classic relief rally plumbing. The tape can look strong, but the underlying fuel is not necessarily fresh belief. It may simply be the mechanical retreat of bearish positioning.Garrett also notes that the broadening out trade keeps coming up on the desk as investors hunt for the next opportunity. That makes sense. Popular AI trades have slipped to the bottom of the short-term performance table, while more defensive and broader strategies have started to work better. But this is where the distinction matters. Broadening out because investors are finding new alpha is bullish. Broadening out because crowded shorts are being covered is more fragile.Under the hood, Goldman’s prime brokerage data points to the same issue. Hedge funds have bought US equity risk for four straight weeks, but Garrett says the latest week was less about adding alpha exposure and more about cutting beta shorts. Short covering outpaced long buying by 4.7 to one. That is the whole market problem in one line. A rally driven by bears being forced out is not the same as a rally driven by bulls stepping in. One can start the move. The other has to sustain it.The second problem is the Fed. Barnert highlights Kevin Warsh’s first FOMC meeting on June 16 to 17 as a potential spoiler, and that is exactly where the peace deal rally can hit a rates wall. Warsh has been openly critical of Fed communication and has already flagged the idea of a regime change. But the market will not trade the phrase. It will trade the tone, the dots, the inflation tolerance and whether the new chair sounds like he is trying to rebuild credibility by leaning harder against financial conditions.The setup is unforgiving. Inflation remains sticky, energy has been a wildcard, and investors are still trying to price whether the Fed may need to raise rates by December. Warsh also has to sound credibly independent while a White House that wanted him in the chair listens carefully to every word. That is a tricky balance. If the market hears hawkish credibility rather than dovish relief, the Iran peace dividend could quickly be overwhelmed by Fed risk.This is the chart that matters most. Gold and the dollar will be watching it like a market signpost, sniffing out whether the first dovish breadcrumbs are starting to appear in the lobby.The third problem is AI policy risk. Barnert points to Washington’s move against Anthropic as a major escalation. The Commerce Department ordered Anthropic to block foreign nationals from its newest models, Claude Fable 5 and Mythos 5. According to the Bloomberg Markets Live note, Anthropic responded by shutting off both platforms for everyone. That matters because this is not the usual chip restriction story. This is Washington moving from the hardware layer into the model layer itself.That changes the AI trade. Until now, the market has treated AI mostly as a capex, compute, chips, power and margin story. Now it also has to be treated as a political access story. The question is no longer simply who has the best model, the most compute or the strongest revenue path. The question is who is allowed to use the model, who is allowed to sell it, and how quickly Washington is willing to turn frontier AI into a controlled strategic asset.That makes the leadership trade harder to price. The AI complex is already crowded, index-heavy, and heavily embedded in market performance. If access to models can be restricted by government directive, the multiple becomes more difficult to defend. This is why the Nasdaq 1996 to 2003 comparison with the SOX since 2022 keeps appearing. The overlay is not perfect, and it depends heavily on where you start the clock. But the reason it resonates is obvious. Investors are asking how much future domination has already been paid for in today’s price. If the answer now depends partly on Washington, the picture becomes much fuzzier.The fourth issue, and arguably the most important after the initial relief trade, is supply. Barnert notes that SpaceX priced at $135 and began trading on June 12 at roughly a $1.77 trillion valuation, making it the largest IPO ever and about three times larger than the previous record. On the surface, that is encouraging. Demand showed up. The market absorbed the deal. The stock traded well.But the real test is not the debut. The real test is whether the market can keep absorbing that much new paper over the coming months, especially with Anthropic and OpenAI still waiting in the wings. SpaceX alone raised more than every US IPO in 2024 and 2025 combined. That is not just a successful listing. It is a capital absorption event.So this is the setup. If the current short-covering hands off to real alpha buying, the broadening-out trade can get legs. Defensive strength, non-AI leadership and fresh capital deployment would allow the market to grind higher. In that version, the Iran relief becomes the first domino to fall the right way.But if conviction remains missing, Warsh leans hawkish, AI leadership becomes a Washington policy trade, and the IPO calendar keeps pulling capital out of the secondary market, then the record wall of supply meets a market that solved the easy problem but is still struggling with the hard one.That is the real message from Barnert, Garrett and Tchir. The peace deal removed the fat tail. It did not remove the market’s structural fragility.Bottom line from The Dark Side of the Boom™: the war premium came out, but the market premium still has to earn its keep.