永续基金指数收益平滑处理分析

Return Smoothing inEvergreen Fund Indexes Institutional Research Group Juan Mier, CFALead Analyst, Fund Strategiesjuan.mier@pitchbook.com PitchBook is a Morningstar company providing the most comprehensive, mostaccurate, and hard-to-find data for professionals doing business in the private markets. Miles OstroffAssociate QuantitativeResearch Analyst pbinstitutionalresearch@pitchbook.com Key takeaways Published on June 8, 2026 •Private markets’ performance is determined by appraisals rather than by themarket-based pricing done in public securities. This is the reason for a long-standing critique of private market returns: Valuations are adjusted too slowly, •Techniques have been developed to adjust for return smoothing in private markets’performance. At PitchBook, we have applied such techniques to private capital •The effect of desmoothing is most pronounced in real estate, where adjustedvolatility is nearly twice as high as reported volatility (5.4% versus 3.0%), and theautocorrelation is statistically valid. The broad private capital composite exhibits •Interpreting return-desmoothing results in evergreen funds must be done carefully,as several evergreen funds are not pure-play private exposures. Continuous inflowsmay lead to irregular periods of elevated cash levels awaiting deployment. The •This analyst note provides one of the industry’s earliest looks at return-smoothing Introduction Performance and risk measurement are ubiquitous and straightforward in publicmarkets, given that asset returns are based on observable market prices. But turn toprivate markets, and this is no longer a simple process. Appraisal- and model-basednet asset value (NAV) calculations, coupled with infrequent valuation exercises, Regardless of NAV pricing frequency, evergreen funds provide exposure to privateassets, so NAV appraisals may lead to the same smoothing biases as in traditionalprivate funds. Evergreen fund returns can also result in misleading risk statistics, This note extends our foundational analysis ofvolatility desmoothing in the privatemarkets, which we published in 2021. We draw from theMorningstar PitchBook USEvergreen Fund Indexes, an industry-first suite of benchmarks that aggregates the Methodology and data Our dataset consists of US direct lending, alternative credit, real estate, and privatecapital evergreen indexes, available in both cap-weighted and equal-weightedversions. Given the recency of the evergreen fund universe, we selected the indexesfrom our suite that had sufficient return histories to support our analysis. Eventually,we will extend this analysis to the full suite of evergreen indexes as their track Although critiques of private markets’ understated volatility persist to this day,methodologies for adjusting appraisal-based returns have existed for decades. In thisnote, we rely on the work of Professor Emeritus David Geltner of the Massachusetts Geltner’s insight is that appraisal-based returns are a smoothed version of theunderlying market returns because commercial property appraisers anchor theirnew estimates on prior-period appraised values. We extend a similar logic to privatefund NAVs. Geltner models the reported return as a weighted average of current The key methodological choice is how to set the smoothing parameter. We assumethat market efficiency holds. Under this belief, underlying returns are not seriallycorrelated, so any observed autocorrelation in the smoothed series comes from thesmoothing itself and not from any persistence or momentum in market returns. For Several other advanced quantitative methods exist for desmoothing private marketreturns.3However, we selected the Geltner method with a one-period lag for itsinterpretability across all four indexes. We will explore alternative desmoothing Results There appears to be a smoothing effect in evergreen indexes. Results vary by strategy,with the Real Estate Evergreen Fund Index showing an outsized effect on risk statisticsafter returns are desmoothed. The effect on the private debt indexes—the Direct For each strategy, we summarize reported statistics versus desmoothed (henceforthreferred to as “adjusted”) statistics for returns, volatility, beta, and correlation relativeto a public benchmark, as well as maximum drawdown. Not to be confused with The indexes’ return series start at different points in 2014 and 2015. This means theCOVID-19 pandemic is the only real crisis event in our data history. Ideally, this analysiswould go through a few cycles of boom and bust to make the results more robust for Real estate Risk parameters for adjusted real estate returns differed considerably from thereported return series. For the period from February 2015 to February 2026, annualizedvolatility in the Real Estate Evergreen Fund Index was almost twice as high in theadjusted series compared with the reported series: 5.4% versus 3.0%, respectively.Higher volatility translates into a steeper maximum drawdown for