Managing Risk Ahead of Binary Events
Market Commentary
The Federal Reserve held rates steady this afternoon, with two members dissenting in favor of a hike. Microsoft and Alphabet report after the close, Amazon follows tomorrow, and the S&P 500 enters the stretch roughly 1% below its October record.
For anyone managing risk, a cluster of binary events like this presents a familiar dilemma. Carrying full exposure through the reports accepts whatever the next 48 hours deliver, while stepping aside leaves the chance forfeiting the bounce entirely. Both choices ultimately rest on a view about a handful of earnings releases, which is a difficult place to anchor a portfolio.
Our process approaches these moments by reading the environment an event lands in: how positioning is distributed, where trend and momentum stand, and how investors have historically behaved when surprised. Those behaviors repeat with remarkable consistency. When they push positioning to an extreme, the distribution of outcomes becomes skewed, and that skew is what we size exposure to.
Today's environment resembles a climber approaching the crux, the most difficult move on a route. Progress upward is earned one hold at a time, while a fall happens in an instant. Experienced climbers still commit to the move, but they place protection first, so a slip costs a few feet rather than the wall.
We view the current wall as one where a fall would likely be faster and deeper than the climb. That view is drawn entirely from positioning, and three observations support it:
Systematic selling is already underway.
CTAs have been reducing exposure and would sell far more in a decline than they would buy in a rally.
Buying capacity tends to return slowly.
Hedge funds hold the most room to add, and have historically rebuilt exposure over weeks to months. That makes reduced exposure relatively inexpensive on the upside
The largest position has the most to sell.
Non-dealer participants hold a record long, concentrated in the Nasdaq.
Systematic Selling is Already Underway
CTAs have reduced US equity exposure every day since October 13, according to J.P. Morgan's positioning desk. Goldman's CTA model has them selling roughly $17 of global equities in a down tape this week for every $1 they would buy in an up tape. Their signals have already turned, and they execute mechanically.
Buying Capacity Tends to Return Slowly
The capacity to carry a rally further sits mostly with hedge funds, whose net leverage is at a five-year low (Goldman Prime). The room to add is there. Historically, it has been deployed gradually. After the April 2025 selloff, the S&P 500 was setting new highs by late June while aggregate positioning was still below normal, and long/short hedge fund beta sat near the bottom of its historical range for months afterward. In 2008, hedge fund leverage fell by nearly half and had barely recovered seven months after the market bottomed.
The Largest Position Has the Most to Sell
Non-dealer participants, largely asset managers, hold a record nominal long in Nasdaq futures, according to the latest CFTC data. They added $9.5 billion in a single week while the index fell 3.5%. Whether that position stays put is a question of conviction. By arithmetic alone, it is the one with the most to sell.
The pressure is concentrated in the Nasdaq. That is where the record long sits, where the trend signals broke, and where tonight's reporters carry the most weight. The S&P 500 feels it through those same companies, which are its largest holdings, but with more support underneath and less distance to fall. The crux is a Nasdaq move. The S&P is roped to it.
Dealer positioning adds a layer. With dealers short gamma between roughly 7,600 and 7,800 in the S&P, hedging flows are likely to amplify the first move in either direction. Our read is that the upside meets resistance close overhead near the record, while the downside has more room before meaningful support.
We saw what a slip looks like on October 22, when one Meta headline about capital spending took the Nasdaq down 2.9% in a single session. Today showed the other side of the same question. Oracle, the most indebted name in the AI buildout, capped its future borrowing and share issuance and put a date on positive cash flow, and the market rewarded it. Same question as Meta, opposite answer, opposite reaction.
Placing Protection
Tonight's results could go either way. What can be observed is how uneven the consequences look. With systematic sellers already in motion and buying capacity likely to return slowly, we size exposure to match: lighter into the event, rebuilt as momentum confirms the next hold.
The trade-off, as we see it: on a good outcome, reduced exposure gives up part of a modest first move and still participates in the slower climb that follows. On a poor one, it absorbs a fraction of a move that could run further and faster.
Different Wall, Same Climbers
Next month's wall will look different. Some setups are steep to the upside, and some crumble slowly. The headlines change too. Today it is AI spending. Next time it will be rates, tariffs, or something nobody is watching yet.
What repeats is how markets climb. Positioning extrapolates, adding the longer a trend runs, as it did straight through October's decline. Reactions overweight whatever just happened. And losses are felt roughly twice as intensely as gains (Kahneman and Tversky), by every participant, so when holds break, positioning tends to unwind all at once.
Those behaviors leave fingerprints we can measure: who holds the most, which way the trend is leaning, and where price sits relative to the levels that matter. When behavior pushes the market to an extreme, the probabilities tilt heavily in one direction. That tilt shows up again and again, whatever the headline.
Why Falls Matter Most
The arithmetic of drawdowns is familiar, but it is worth restating. In our study of the S&P 500 from 1995 to 2014, full investment returned 9.85% a year. Missing the 10 best days cut that to 6.10%. Avoiding the 10 worst lifted it to 14.13%.
Deep drawdowns also carry costs the return figure misses. They consume risk budgets, prompt rebalancing and governance reviews, and can force de-risking near the lows. Avoiding the deepest falls is what preserves the capacity to stay on the wall through the full climb.
The Next Hold
Tonight's result is unknowable from here. How markets respond to it follows patterns that recur cycle after cycle. We build our process around that distinction: participating when the holds are good, placing protection before the crux, and sizing each move to the shape of the wall.
Subscribe to receive timely updates and in-depth insights.
Related Insights
