Global Equity Volatility Insights

Frequently asked questions about the BofA Bubble Risk Indicator (BRI)

In our 2026 Year Ahead report, we introduced the BofA Bubble Risk Indicator (BRI), a price-based measure designed to detect bubble-like instability across financial assets. Given strong client interest, we are now adding a weekly one-page BRI update to the GEVI (see earlier pages). This update will track BRI levels across major markets and asset classes, highlight pockets of equities exhibiting the most froth, and call out recent increases in bubble-like price action. Below, we address some of the most frequently asked questions about the BRI.

The BofA Bubble Risk Indicator is intended to be an indicative metric only and may not be used for reference purposes or as a measure of performance for any financial instrument or contract, or otherwise relied upon by third parties for any other purpose, without the prior written consent of BofA Global Research. The BRI was not created to act as a benchmark.

1. How does the BRI quantify ‘bubble risk’?

A defining feature of historical asset bubbles is that volatility tends to rise alongside prices. Building on this observation and drawing inspiration from how the four statistical moments characterize a distribution, the BRI quantifies bubble risk using the first four moments of an asset’s price action: (i) returns, (ii) realized volatility, (iii) price momentum, and (iv) fragility/convexity.

These four moments typically become elevated during the frothiest phases of asset bubbles, reflecting the influence of exuberant sentiment, one-sided positioning and imbalanced supply-demand dynamics on price action. The late 90s dotcom bubble in the Nasdaq provides a clear and pertinent illustration of these dynamics (Exhibit 30 & Exhibit 31).

2. How is the BRI constructed?

The BRI is constructed as an average of three sub-indicators: a short-term, medium-term, and long-term measure. Each sub-indicator ranges from 0 to 1. Using multiple time horizons allows the indicator to capture bubble-risk dynamics that can develop at different speeds.

Each sub-indicator aims to capture how stretched an asset’s four moments (returns, vol, momentum, and fragility) are relative to the asset’s own history, and is computed as the average of each of these four moments’ percentile ranks vs history. However, given detection of bubble-like behaviour is the primary aim of the indicator, we condition this average percentile rank on periods when the asset is rallying by scaling it up or down depending on how positive or negative returns are.

Exhibit 32 summarizes the construction of the aggregate indicator and its microstructure.

3. What is a ‘high’ BRI reading and what are its implications?

Since the BRI is defined on a 0 to 1 scale, assets with readings approaching 1 are exhibiting increasingly bubble-like price action. How close to 1? While it is difficult to specify a universal threshold given the plethora of asset-specific dynamics that can present in markets, we find that BRI levels above 0.8 (i.e., top quintile) provide a useful rule of thumb beyond which bubble risks are likely elevated.

Tactically, high BRI readings tend to signal asymmetric downside risk in the near-term (higher risk of the bubble ‘popping’), which we demonstrate empirically later in this section. However, rather than treating the BRI as an explicit sell signal (“get out”), which can risk major underperformance in a bubble, we find the BRI best positioned as a cautionary tool to add asymmetric protection/equity replacement (“watch out”).

4. What assets can the BRI be computed on?

Given the BRI relies only on price-based inputs, it can be computed on any asset with a sufficient history of daily price data. Indeed, by inferring sentiment and positioning from price action, the BRI can be applied to assets without explicit fundamental valuation methodologies (e.g., commodities) or those with opaque positioning visibility.

5. What has the BRI looked like in past asset bubbles?

Exhibit 33 demonstrates what the BRI framework would have looked like when applied to eight historically well-recognized asset bubbles. As these bubbles formed, increasingly stretched returns, vol, momentum, and fragility pushed BRI readings towards 1. Indeed, seven of the eight bubbles saw their peaks concurrent to a BRI reading above 0.8.

The forward distribution of returns during these bubbles conditional on the BRI would have shown higher near-term downside risk after the BRI breached 0.8 (Exhibit 34), with broadly similar upside potential (Exhibit 35).

6. Can an asset with a high BRI reading not be in a bubble?

Yes. One of the risks stemming from the BRI’s exclusive use of price data is instances of ‘false positives’ when the indicator appears stressed, but the asset doesn’t appear to be in a bubble based on other metrics. For example, semiconductor stocks exhibited frothy behaviour in the immediate aftermath of ChatGPT’s release in Nov-22 but also saw a fundamental shift and expansion of their business model in the AI era.

Another scenario of a high BRI level that may not point to a bubble is in the immediate aftermath of an exogenous shock (e.g., COVID), where vol/fragility remain high and prices are retracing their drawdown. In such cases, the interpretation of a high BRI level requires additional context.

7. Can an asset in a bubble have a low BRI reading?

Yes. A related limitation of the BRI’s reliance on price data is the implicit assumption that the asset bubble in question would result in historically elevated instability. While the typical evolution of a bubble sees this happen as stretched positioning and exuberance translate into volatile & fragile rallies with high momentum into the peak, outlier cases where the bubble peak sees the bull run end in a whimper can occur and are a challenge for the BRI (e.g., the Nikkei bubble in the 1980s, which saw its most unstable rally earlier rather than later).

8. How can one use the short-term to long-term sub-indicators of the BRI?

As discussed above, the BRI aggregates short-term, medium-term, and long-term sub-indicators to capture asset price dynamics across multiple horizons. In addition to examining the sub-indicators to understand shorter vs longer-term trends, we also find the range across sub-indicators provides a confidence level on the BRI reading, with tighter ranges providing higher confidence. For example, a 0.7 BRI reading is likely more informative when the sub-indicators span 0.6 to 0.8 than when they range from 0.4 to 1.0, where disagreement across horizons suggests greater uncertainty.

9. What other indicators can help confirm elevated bubble risk?

At its core, the BRI is simply a parsimonious transformation of prices, avoiding a “kitchen-sink” approach that incorporates explicit positioning, leverage, or PB data. This keeps the framework robust to overfitting and makes it applicable across assets, regions, and historical periods where such data are often unavailable or unreliable.

Despite the appeal of such an approach, we find the BRI most effective when used in conjunction with complementary tools that can help assess whether frothy price action is supported by fundamentals - for example, more traditional valuation (e.g., PV of cash flows, valuation multiples, etc.) and/or positioning metrics (e.g., leverage, inflows, retail trading volumes etc.) where available.

10. What universe of assets will be covered in the BofA BRI Landscape?

We plan to include BRI levels on a broad set of assets covering global equity indices, large cap US equity sector indices (including the “Magnificent 7”), major commodities (including the BCOM Index), and major cryptocurrencies (Bitcoin and Ethereum). In addition to this core cross-asset universe, we plan to apply the BRI framework to popular equity themes that have experienced significant inflows over the prior week. This thematic lens is designed to spot emerging pockets of froth that may not yet be visible at the broader index or sector level.