Skip to main content

Positioning & Fund-Manager Surveys (BofA FMS, COT)

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,321 words

Positioning data answers a different question than valuation or momentum: who has already acted, and how much firepower is left? Two of the most-watched gauges are the Bank of America Global Fund Manager Survey (FMS) — a monthly opinion-and-allocation poll of institutional investors — and the CFTC Commitments of Traders (COT) report — a weekly, regulator-mandated census of actual futures positions. Both are read primarily as contrarian tools: the working assumption is that when a crowd of "smart" money is leaning hard one way, the marginal buyer (or seller) is exhausted and the trade is vulnerable to a reversal. The central tension is that this logic is intuitive and occasionally spectacular, but the data is lagged, noisy, and far weaker as a standalone timing signal than its folklore suggests.

How it's formed

BofA Global Fund Manager Survey. Conducted monthly since the mid-1990s, the FMS polls roughly 200–400 institutional investors (hedge funds, pension funds, mutual funds, asset managers) collectively running hundreds of billions of dollars, per BofA's own description and secondary reporting. Respondents report average cash allocation, equity/bond/region/sector over- and under-weights, and macro expectations (growth, inflation, rate path). BofA distills these into composite gauges, most famously the Bull & Bear Indicator (a 0–10 scale blending positioning, flows, and credit/breadth signals) and the "Cash Rule."

The Cash Rule, introduced around 2002, is a mechanical contrarian signal on average cash levels: a buy trigger when cash rises above ~4.5% (fear/de-risking) and a sell trigger when cash falls below ~4%, with heightened concern below ~3.7–3.8% (complacency). These thresholds are BofA's published rule and are widely re-quoted in financial press.

CFTC Commitments of Traders. A weekly report covering Tuesday open interest, released the following Friday at 3:30 p.m. ET, for every U.S. futures market where 20+ traders hold reportable positions. The Legacy report splits traders into commercials (hedgers — producers, processors, banks with offsetting cash exposure, often labelled "smart money"), non-commercials (large speculators — funds, CTAs), and non-reportables (small traders). The newer Disaggregated report refines this into producer/merchant, swap dealers, managed money, and other reportables; the Traders in Financial Futures (TFF) report does the equivalent for financial contracts. Analysts chart net position (longs minus shorts) per category.

Because raw net numbers grow with open interest, practitioners normalize them. The best-known transform is the COT Index (Larry Williams' "Williams Commercial Index"/WillCo): rescale a category's net position to 0–100 via (net − min)/(max − min) over a look-back, where 0 = most-short-ever and 100 = most-long-ever in that window. Williams' own books used a 26-week look-back; many modern implementations default to 156 weeks / 3 years (e.g., COT Unchained). Both are in common use. Z-scores and percentiles serve the same purpose — flagging readings above the ~90th or below the ~10th percentile (or beyond ±2 standard deviations) as extremes.

How it's used in practice

Both feed the same workflow: gauge crowding, then fade extremes — at extremes only. FMS cash and the Bull & Bear Indicator are macro/equity-allocation overlays; a multi-decade-low cash reading or a Bull & Bear print near 8–10 is read as a contrarian sell, near 0–2 as a contrarian buy. BofA reports that historically very low cash has on average preceded modestly negative forward equity returns, though the precise figures vary by sample and should be treated as illustrative rather than a tradeable constant.

COT is used most heavily in commodities and FX, where commercial hedgers have genuine informational advantage. The classic Williams playbook: when commercials reach a multi-year net-long extreme (COT Index near 100) while speculators are net-short, expect a bottom; the reverse for tops. Almost all serious practitioners use COT as confirmation/context alongside price (trend, support/resistance) rather than as a trigger, because positioning can stay extreme for months.

Adoption, debate & evidence

The FMS is one of the most-cited sentiment reports in markets; its monthly release reliably moves financial headlines. COT is a free, regulator-sourced dataset with a large retail and institutional following. So adoption is high — but efficacy is genuinely contested.

For COT, there is real academic support for the underlying mechanism: the hedging-pressure hypothesis (Bessembinder 1992; Leuthold et al. 1994; de Roon, Nijman & Veld 2000) finds that hedger positioning helps explain futures risk premia and returns across markets. That is a robust factor finding. It does not translate cleanly into a profitable mechanical timing rule. Practitioner backtests (e.g., QuantifiedStrategies, EarnForex spanning ~30+ rule variants) report a positive but inconsistent edge — profitable on many currency pairs but with large drawdowns, and outright failure on others (AUD/USD, and unreliable results on gold and emerging-market crosses). The signal is also structurally lagged (Tuesday data, Friday release) and noisy, and category labels can mislead — "commercial" swap dealers may be hedging speculative client flow, blurring the smart-money story.

For the FMS Cash Rule, the threshold values are well-documented but the sample is short (the rule dates to ~2002, giving relatively few independent signals), so any hit-rate claim rests on a handful of episodes and invites overfitting. Survey data also captures stated sentiment, which can diverge from actual books. Treat any precise "X% accuracy" claim about either tool as unverified marketing unless it names a peer-reviewed source.

Honest summary: the crowding concept is sound and academically supported; the specific published rules are weak-to-modest standalone signals best used as context, not triggers.

Strengths & limitations

Works best at genuine multi-year extremes, in markets with real hedgers (commodities, FX, rates), and as a confluence factor that raises conviction when it agrees with price and macro. It is most valuable for spotting asymmetry — when positioning is so one-sided that the pain trade is obvious.

Fails as a precise timing tool. The #1 misuse is fading a trend on a "high" reading that is not actually extreme — positioning can grind more extreme for months during strong trends, and "overbought positioning" is not a short signal by itself. Other traps: the reporting lag, conflating stated-survey sentiment with real exposure, applying COT to equity indices where the "commercial" category is dominated by index arbitrage rather than informed hedging, and reading raw nets without normalizing for open-interest growth.

Sources

Disputes flagged: (1) BofA Cash Rule thresholds are well-sourced but its hit rate rests on a short post-2002 sample — no peer-reviewed validation found. (2) COT's hedging-pressure factor is academically robust, but mechanical COT timing rules show only a modest, inconsistent backtested edge with large drawdowns. Do not let the academic factor lend credibility to the retail timing rule.