Committing more after volatility spikes beat fixed pacing in QRG vintage study

The instinct when markets turn ugly is to stop writing new checks. An analysis published by Insurance AUM Journal, built on QRG research and Burgiss data, argues that in private markets that instinct has been expensive: across vintages from 2000 through 2020, a program that increased its annual commitments after periods of elevated volatility produced higher net IRRs, higher TVPIs and stronger alpha than one that committed a fixed dollar amount every year, with the result holding across major asset classes and across North America, Europe and Asia Pacific.

The analysis arrives at a moment when caution is easy to defend, its authors pointing to inflated valuations, macroeconomic risk, geopolitical conflict and the possibility of an AI-driven market bubble as forces that have pushed uncertainty back to the center of allocation decisions in 2026; the question they pose is not whether investors feel more cautious but whether that caution should lead them to commit less capital.

For a general account, the interesting part is where the paper locates the decision: not in asset selection or manager selection, but in the size of the annual commitment, one number a long-duration investor sets directly and can set by rule rather than by budget. An insurer never chooses the vintage it gets; it chooses how much to put to work in each one, and the paper's claim is that the level of volatility at the moment of commitment, treated as a proxy for entry conditions, separates the better vintages from the worse ones.

The underlying idea is older than the result: Kaplan and Schoar's 2005 study found that funds raised during periods of heavy fundraising tend to underperform funds raised in less crowded periods, and that fundraising is procyclical, rising during stronger economic environments and falling when uncertainty increases. Those authors also observed that strong public-market performance and low volatility often precede higher fundraising, which leaves the new paper testing the other side of that relationship: whether elevated volatility at entry can serve as a forward-looking indicator of stronger performance.

QRG's own 2023 work on commitment timing landed in a similar place, finding that deviations from long-term fundraising trends have been associated with future private-market returns. Put the two together and a mechanism appears: if low volatility and strong equity markets draw capital into funds, and heavily funded vintages underperform, then the conditions that make a commitment feel comfortable are the ones the evidence associates with weaker vintages. Read that way, the rule is a claim about the vintage rather than about the price of the assets inside it, a distinction the published excerpt does not draw but one that bears on what the result means.

To build the reading, the analysis takes the VIX, described in the paper as the equity market's expectation of near-term volatility, as a proxy for total volatility; for each vintage year it ranks the prior year-end VIX against the previous ten years of monthly observations, 120 data points in all, which yields a rolling ten-year volatility percentile. That percentile then scales the commitment, with a fixed strategy committing $100 every year and the pro-volatility version committing more when the prior year-end reading is high and less when it is low. The published excerpt does not report the resulting schedule, the spread between the largest and smallest annual commitment, or how many vintages fell into each bucket.

Two features of the test are worth weighing before anyone rewrites a commitment schedule: the window ends with the 2020 vintage, so the vintages formed since then sit outside the evidence, and what the excerpt reports is the direction of the effect rather than its size. And the proxy is an equity-market one, leaving open whether an equity-implied volatility number sorts private-market vintages as well as a credit spread or a measure of deal flow would; its performance data come from Burgiss, per the paper's own source note.

The paper's own verbs are hedged: the analysis suggests the relationship rather than establishing it, and the authors frame leaning into volatility as something that may be useful for long-term private-market investors rather than as a rule for everyone. That is roughly the right level of confidence for a single study covering a twenty-one-vintage window, and nothing in the published excerpt is written for insurance buyers in particular.

And timing is not the only question a general account has to settle. As this publication has argued, the NAIC has moved from reviewing private credit to pricing it, through RBC preamble changes, narrowed gap lists, and a $1.2 trillion private-credit perimeter letter to Senator Warren, repricing a ceded book on a timetable set by regulators and rating agencies rather than by markets. A volatility percentile appears in none of that machinery, and an insurer that leaned into a high-volatility year on QRG's rule would still hold the resulting exposure under whatever capital treatment applies later, while the entry condition the paper measures and the charge the NAIC assigns are set on different calendars.

The harder obstacle for a regulated buyer is governance: the percentile is fixed at each prior year-end, before the year it governs, which is what makes it testable and also what makes it difficult to follow, since the instruction to commit more arrives with the drawdown that produced it, in the year a committee is most likely to want to commit less. Whether a general account can hold the vintage long enough for the entry condition to matter is a question about its own liabilities rather than about the VIX.

What the paper cannot settle is whether the rule travels past 2020: the vintages formed since then sit outside its window, and the ones being committed to now will be judged by the same crowding logic the 2005 study described. The fundraising-deviation measure from QRG's 2023 work is the one to hold against those commitments, since it is built on the entry condition the newer paper is trying to replace with a volatility reading.

For a general account setting next year's commitment, the paper offers a rule, a rationale and a twenty-one-vintage record. What it does not offer is the magnitudes behind the result, and the vintages a live rule would govern are the ones its record cannot yet reach.