When Belief Becomes ObligationS&P 500 IndexTVC:SPXAlphaPineWhen Belief Becomes Obligation Part 2 of 2: How expectations become positions, positions become constraints, and constraints become market-moving flows Part 1 started with a simple idea: A level can matter because enough participants behave as if it matters. But that is only the beginning of the feedback loop. Once a belief creates a position, the next trade may no longer depend on belief at all. An option position can create hedging needs. An index mandate requires rebalancing. A systematic strategy generates orders when its rules are triggered. Leverage can force liquidation. The original expectation has now created something more durable: an obligation to transact. That gives us the central idea of Part 2: A belief can eventually create a structure that no longer requires belief to keep operating. FROM BELIEF TO STRUCTURE Consider the sequence: Expectation → Position → Exposure / Constraint → Systematic or Mandatory Flow → Price → New Expectations The person executing the final trade in this chain may have no opinion about the market whatsoever. A dealer rebalancing delta is not necessarily bullish. An index tracker buying a newly included stock is not necessarily optimistic about the company. A fund reducing risk after a volatility shock may still believe the asset is cheap. Their orders happen because something upstream has constrained their choices. In the strict Mertonian sense, not every feedback loop described here is a self-fulfilling prophecy. Some are better thought of as mechanical feedback generated inside the market itself. The label matters less than the mechanism. A belief can create a position. A position can create exposures or constraints that generate future obligations. Those obligations can then amplify, dampen or reshape the price behaviour that participants later observe. Behaviour can crystallise into market structure. Once that happens, the mechanism can outlive the belief that helped create it. Market feedback can begin with discretionary belief and later become mechanical. Once positions create constraints, subsequent orders may occur regardless of whether the trader generating them still believes in the original idea. 1. THE DEALER: WHEN A POSITION CREATES A HEDGE Options provide one of the clearest examples. When an option market maker takes the other side of customer flow, the resulting portfolio can carry directional and convexity exposure. Dealers commonly hedge part of that exposure in the underlying market. As price moves, the required hedge changes. That creates a feedback link between: Option Position → Underlying Price → Hedge Requirement → Underlying Trading But the direction matters. If the relevant dealer position is long gamma, hedge rebalancing tends to work against the price move: Price rises → hedge requires selling. Price falls → hedge requires buying. That can dampen movement. If dealers are short gamma, the feedback reverses: Price rises → hedging can require buying. Price falls → hedging can require selling. That can amplify movement. So an options strike is not automatically a "magnet", and dealer hedging does not automatically produce mean reversion. The effect depends on the sign, size and distribution of aggregate exposure. This is not merely theoretical. Ni, Pearson and Poteshman documented that on option expiration dates, optionable-stock closing prices cluster around strike prices more often than would be expected by chance. Their 2005 study estimated that option-related effects altered returns by at least 16.5 basis points on average per expiration date, corresponding to aggregate market-capitalisation shifts on the order of US$9 billion in their sample. Later work by Ni, Pearson, Poteshman and White found a broader relationship between option-market hedge rebalancing and underlying stock volatility. Their estimates suggested that roughly 12% of daily absolute returns in optioned stocks could be associated with hedge rebalancing, with effects present outside expiration week as well. The point is not that options "cause support and resistance." The stronger conclusion is narrower and more useful: Derivative positions can create mechanical flows in the underlying market that have measurable effects on price behaviour. This matters even more in markets where short-dated options dominate activity. In 2025, SPX zero-day-to-expiry options averaged around 2.3 million contracts per day and accounted for roughly 59% of SPX options volume. In February 2026, Cboe reported a new record: 2.99 million contracts per day, or 63% of all SPX trading. 0DTE activity remained elevated through the second quarter of 2026. That does not tell you tomorrow's direction. It tells you that ignoring the interaction between derivative exposure and underlying order flow can mean ignoring an important part of the market's structure. This chart does not prove that hedging caused the price behaviour shown. It illustrates a pattern that is consistent with the positioning and hedging dynamics discussed above. Causation cannot be established from price data alone. 2. WHEN MANY SYSTEMS REACT TO SIMILAR INFORMATION Not every mechanical flow comes from derivatives. Systematic strategies create another form of coordination. The scale is meaningful. Industry estimates place total managed-futures assets at roughly US$340 billion. That is an industry estimate rather than a peer-reviewed measurement, but it is enough to show that rule-driven capital in futures markets is not marginal. Trend followers may use moving-average relationships, breakout rules, time-series momentum, volatility scaling or combinations of several horizons. The exact models differ, and there is no evidence that "the MA200" is some universal institutional trigger. But many systems can still respond to related information. If several independent models detect a strengthening trend at roughly the same time, their orders can become correlated even though the managers never coordinated. Nobody needs to call anyone. Nobody needs to agree. Similar data plus similar objectives can be enough. Now add risk scaling. A market falls sharply. Volatility rises. Target exposure falls. Positions are reduced. That selling can add to the move, which can increase volatility again. Another feedback loop appears. The important distinction from Part 1 is that the model does not need to "believe" anything. Discretion has been delegated to a rule. 3. INDEX FUNDS: THE PERFECT EXAMPLE OF OBLIGATION — AND ADAPTATION Index changes give us an unusually clean experiment. When a security enters a major benchmark, passive funds tracking that benchmark must eventually own it in the required weight. That creates a predictable demand shock. For decades, stocks added to the S&P 500 displayed large positive abnormal returns around inclusion. Robin Greenwood and Marco Sammon examined how that effect changed over time. Their published results show the average abnormal return associated with S&P 500 additions was around 7.4% in the 1990s. Over the most recent decade in their sample, the effect fell to less than 1%. The important part is what did not disappear. Indexing did not disappear. The obligation to track the index did not disappear. The amount of capital linked to the index did not collapse. The flow remained. The easy profit in anticipating the flow faded. Why? Markets adapted. Other indices generate offsetting flows. Liquidity provision improved. Participants became better at anticipating index changes. Arbitrageurs could accumulate inventory before forced buyers arrived and supply shares when those buyers needed them. But adaptation is not necessarily the end of the story. Secondary reports citing Goldman Sachs research indicate that the inclusion effect has partially returned in recent years, with 2025 additions outperforming the equal-weighted S&P 500 by roughly 7.4 percentage points on announcement day — a revival those reports attribute largely to increased retail participation in a handful of high-profile names. That figure comes from reporting on a bank research note rather than a peer-reviewed study, so treat it as indicative rather than definitive. The conceptual implication is what matters here. An edge that appeared largely competed away can re-emerge when participant composition changes. This gives us perhaps the cleanest version of the paradox: A mechanism can remain economically real while the trading edge associated with forecasting it disappears — and market adaptation is itself conditional on who is participating. Adaptation does not necessarily erase the phenomenon. It changes who captures the opportunity, and that allocation can shift again. The forced flow did not disappear when the historical inclusion premium shrank. What changed was the opportunity available to traders anticipating it. More recent evidence also suggests that adaptation itself may change when the participant mix changes. 4. WHEN SOPHISTICATED STRATEGIES BECOME THE CROWD The danger of crowding is not unique to retail traders. August 2007 provided one of the clearest examples. During the week of August 6, a number of quantitative long/short equity funds experienced extraordinary losses even though broad market indices did not experience an equivalent collapse. Khandani and Lo investigated the episode and found evidence consistent with deleveraging across similarly constructed portfolios. The important point was not that every fund used identical code. They did not need to. Different teams could use similar academic factors, optimisation methods, risk controls, liquidity assumptions and data. Independent research could therefore converge on correlated positions. To convey the scale: in their simulation, the three-day cumulative loss of roughly −6.85% corresponded to about 12 daily standard deviations relative to the strategy's own historical volatility, while the subsequent rebound day represented roughly 11.4 daily standard deviations in the opposite direction. Those "sigma" descriptions should not be read as literal probabilities from a normal distribution. Financial returns are not normally distributed in the tails. The numbers simply show how extraordinary the episode was relative to the strategies' own historical behaviour. When losses began, deleveraging by one participant could move prices against others holding similar portfolios. Their risk rules then required reductions, and those reductions moved prices further. The crowd was not a thousand inexperienced traders drawing the same trendline. It was sophisticated capital independently arriving at similar portfolios. Coordination does not require communication. Sometimes competence itself converges. 5. THE DISCRETION SPECTRUM Part 1 focused heavily on belief. But market orders exist on a spectrum of discretion. At one end: High Discretion A trader sees a chart and chooses whether to act. Then: Discretionary Institutional A portfolio manager acts within mandates but retains significant judgment. Then: Systematic The design was discretionary, but day-to-day decisions are delegated to rules. Then: Mandated / Constraint-Driven Index tracking, hedge rebalancing, collateral requirements, margin calls and certain risk limits leave progressively less freedom over whether an order must occur. Low Discretion This is not a ranking of price impact. A small mandatory trade can matter less than a huge discretionary one. A concentrated retail flow in an illiquid asset can overwhelm institutional flows. Impact depends on size, urgency, liquidity and market conditions. The spectrum answers a different question: How much freedom does the participant have not to trade? The less discretionary the flow, the less relevant the participant's current opinion becomes. This spectrum measures freedom of choice, not price impact. Mandatory does not automatically mean larger. It means the order is less dependent on the participant's current market opinion. 6. WHEN BEHAVIOUR BECOMES STRUCTURE Now return to the round number from Part 1. Why might a round price continue to matter decade after decade? If popularity necessarily destroyed the underlying phenomenon, persistent price clustering around round numbers would be hard to explain. Yet such clustering has been documented repeatedly across different markets and different decades. The missing distinction is between mechanism and edge. Human preference for simple numbers can influence where orders are placed. Over time, those repeated preferences may coincide with exchange conventions, derivative strikes, liquidity habits, execution logic and market-making behaviour. The behaviour becomes partially embedded in the infrastructure. Future participants no longer need to think: "100,000 is psychologically important." They may simply respond to liquidity already sitting there, derivative exposure already concentrated there, risk models referencing nearby prices, or other participants already reacting to the area. Behaviour can crystallise into market structure. And once crystallised: A belief can create a structure that no longer requires belief to keep operating. But this does not mean every round number or technical level is secretly driven by institutional mechanics. The claim is narrower: Shared behavioural conventions can sometimes become embedded in structures that subsequently generate their own flows. That is a hypothesis worth testing, not a universal explanation. 7. THE PARADOX RETURNS Now the two articles connect. Part 1 showed that: Behavioural reliability and trading EV are different things. Part 2 shows why that distinction survives even when large institutional flows are involved. A mechanism can become larger, more regular, more widely understood and more deeply embedded while the profit available from anticipating it becomes smaller. Index inclusion remains the cleanest example. Forced buying still occurs. Competitive markets learned to prepare for it, and the announcement-day premium largely disappeared. But as the recent partial revival suggests, that preparation is not necessarily a permanent state of the world. It reflects who is currently participating and how well adapted they are. Change the participant mix and the same mechanical flow may once again produce visible price effects. Option hedging provides another perspective. The mechanism is not disappearing; in some markets its footprint may be becoming more important as derivative activity grows. But knowing that gamma hedging exists is not automatically an edge. Everyone else can know too. So the full adoption cycle is not: Popularity → Effect → Effect Disappears It is closer to: Discovery → Adoption → Structural Embedding → Predictability → Adaptation And crucially: Adaptation does not necessarily erase the mechanism. It erases — or redistributes — the easy profit available from anticipating it. The flow can remain. What disappears is your advantage in arriving before everyone else. Adaptation is also not always permanent. When the composition of participants changes, an edge that was competed away can partially return. But revival is not guaranteed, and many anomalies never come back. WHAT THIS CHANGES FOR A TRADER First, stop treating every technical reaction as evidence that the line itself has predictive power. The chart shows the outcome; it does not directly reveal the cause. Second, ask what kind of participant might be active around the level. Are traders choosing to transact? Are systematic models likely to respond? Is there meaningful derivative positioning? Is there a scheduled rebalance or expiration? Are risk constraints likely to matter? Third, separate reliability from payoff. A setup that works 75% of the time can still be worse than one that works 45% of the time if the first setup has been competed into terrible reward-to-risk. Fourth, treat obviousness as information, not automatically confirmation or rejection. A famous level can attract more genuine flow, but it can also attract more competition. Both can be true at the same time. Finally, look for the intersection between discretionary attention and non-discretionary flow. Some of the most consequential market zones may occur where what traders voluntarily watch overlaps with where other participants have structural reasons to transact. Not because those zones are magical. Because several different mechanisms are converging on the same price. THE FINAL POINT Part 1 began with a line on a chart. Part 2 ends with a market structure. Between them sits the full feedback loop: Belief → Position → Constraint → Flow → Price → New Belief Somewhere along that chain, choice can become obligation. Somewhere later, obligation can become predictable. And once predictable, competition can remove the easy edge without removing the underlying mechanism. So the deepest lesson is not: "Technical analysis works because everybody believes in it." Nor is it: "Institutions secretly create every support and resistance level." The deeper idea is this: Beliefs can create positions. Positions can create constraints. Constraints can generate flows. Those flows may remain economically real long after the easy edge in anticipating them has been competed away. And if the participant mix changes, part of that edge may return — or it may not. Part 1 ended with a distinction: Market impact and trading edge are not the same thing. Part 2 ends with the same distinction, stated in its final form: The mechanism and the opportunity are not the same thing. Sources and Further Reading Merton, R. K. — work on the self-fulfilling prophecy. Keynes, J. M. (1936) — The General Theory of Employment, Interest and Money, Chapter 12. Soros, G. — writings on reflexivity and fallibility in financial markets. Ni, S. X., Pearson, N. D. & Poteshman, A. M. (2005) — "Stock Price Clustering on Option Expiration Dates", Journal of Financial Economics, 78(1), 49–87. Ni, S. X., Pearson, N. D., Poteshman, A. M. & White, J. (2021) — "Does Option Trading Have a Pervasive Impact on Underlying Stock Prices?", Review of Financial Studies, 34(4), 1952–1986. Khandani, A. E. & Lo, A. W. (2011) — "What Happened to the Quants in August 2007? Evidence from Factors and Transactions Data", Journal of Financial Markets, 14(1), 1–46. Greenwood, R. & Sammon, M. (2025) — "The Disappearing Index Effect", Journal of Finance, 80(2), 657–698. Cboe Global Markets — SPX and 0DTE options volume statistics. Managed-futures industry asset estimates — industry sources based on BarclayHedge data. Index-inclusion revival figures — secondary reporting citing Goldman Sachs Global Investment Research. This article is for educational purposes and is not financial advice. Empirical figures refer to the cited research samples and should not be interpreted as universal estimates for all securities, periods or market regimes.