THE COMPRESSION PARADOX: LESSONS FROM A LOW-VOLATILITY YEAR
The year 2023 proved to be exceptional once again. If there was one defining element, it would be this: 2023 was the year when equity market volatility decoupled entirely from macroeconomic volatility. [Figure 1] illustrates this divergence clearly, comparing macroeconomic volatility—represented by the realized 250-day volatility of US 2-year Treasury yields—and equity volatility, shown via the realized volatilities of the SPX, SX5E, and NKY indices. While macro volatility soared to levels seen during the Global Financial Crisis, equity volatility plummeted to near-record lows. This development was unexpected. In the first part of this report, we explore the drivers behind this disconnect, and in the second part, we apply the insights gained to navigate the challenges and opportunities of 2024.
SECTION 1 : REVIEW OF 2023
1A. EQUITY VOLATILITY - FROM NORMAL TO SUBDUED
At the start of 2023, equity volatility stood within a typical range. As the months unfolded, however, it declined rapidly into a regime of ultra-low volatility. A key measure we tracked throughout the year was the SX5E Index’s quarterly realized volatility. As shown in [Figure 2], the readings for Q2 and Q4 closely match the subdued volatility levels seen in 2017 (illustrated in blue), often regarded as a historical benchmark for minimal market fluctuations
The drop in realized volatility was extensive and consistent across major global equity indices. The 250-day average realized volatility for the SPX, SX5E, and NKY continuously declined throughout the year. By December, this average had reached levels similar to those in 2017 and 2019 [Figure 3]. Supporting this, only two trading days in 2023 saw the S&P 500 shift by more than 2%—a hallmark of tranquil market conditions [Figure 4].
1B. CREDIT RESILIENCE DESPITE TIGHTENING
Throughout 2023, credit markets demonstrated significant strength—even in the face of an extended and aggressive global tightening of monetary policy. By year-end, US investment-grade and high-yield spreads had tightened to levels close to their long-term lows [Figures 5 and 6].
Even accounting for the typical delay in monetary policy transmission, the robustness of US credit was remarkable. Over the year, the number of credit upgrades outpaced downgrades by a factor of four [Figure 7].
European credit markets also fared well. While spreads for EUR-denominated investment-grade credit remained somewhat wider than those in the US, 2023 recorded one of the lowest instances of “Fallen Angels” (investment-grade issuers downgraded to high yield) since 2000 [Figure 8]—a testament to overall market stability.
1C. BONDS AND EQUITIES – CORRELATION STAYS POSITIVE
In 2023, the correlation between equities and bonds remained positively aligned on a low-frequency basis. Following the challenges of 2022—when both asset classes suffered and diversification failed—many questioned whether such positive correlation was an anomaly. Our position, however, has consistently been that annual return-based positive correlation is more common than generally perceived.
This favorable relationship re-emerged in 2023, especially benefiting long-beta portfolios. It appeared even more prominent in the second half of the year [Figure 9]. From a broader historical standpoint, the correlation between bonds and equities has been trending upward since early 2022 [Figure 10], challenging the traditional notion that bonds always serve as a safe counterbalance to equities.
SECTION 2: OUTLOOK FOR 2024
2A. OUTLOOK FOR EQUITY VOLATILITY
We are convinced that one of the primary reasons for the persistently low levels of equity volatility — both implied and realised — is the significant increase in the supply of volatility. A surge in volatility supply exerts downward pressure on implied volatility because, when the availability of options exceeds demand, option prices naturally decline. Since implied volatility is embedded within option pricing, it too is marked lower as a result.
Moreover, an abundant supply of volatility also suppresses realised volatility through the hedging behavior of option buyers. To generate profits, volatility buyers must outperform the cost of the premium they pay for options, which in practice requires active delta hedging — buying the underlying asset when markets fall and selling it when they rise, a process known as gamma hedging. When the volume of gamma in the market is large, these hedging flows can dampen price swings and keep markets trading within a narrower range.
This dynamic has become more pronounced in recent years. Between May 2022 and December 2023, the assets under management of the three largest short index volatility ETFs nearly tripled, rising from USD 16 billion to USD 47 billion. According to Goldman Sachs, these ETFs are collectively responsible for selling the equivalent of USD 200 million of gamma and USD 2 million of vega into the market each day. [Figure 11]
Although these ETFs primarily focus on short-dated option selling, it is noteworthy that the structured product market remains largely dominated by instruments that supply volatility to the market, rather than those that absorb it. [Figure 12] This outcome is somewhat unexpected, given the significant shift in the risk-reward profile of these products over the past two years. The rise in interest rates has made Capital Protected products more appealing, as they benefit from higher yields, while the decline in equity volatility has reduced the attractiveness of Capital at Risk structures such as autocallables, whose payoffs are less compelling in low-vol environments.
These two flows — short-dated option selling via ETFs and the continued issuance of volatility-supplying structured products — are among the most quantifiable contributors to the substantial supply of single stock and index volatility in the market. However, they are far from the only ones. Additional sources of Vega supply include Quantitative Investment Strategies (QIS) and Alternative Risk Premia (ARP) strategies, both of which play a significant role in reinforcing this imbalance.
For ADAPT, this abundant volatility supply presents both challenges and opportunities across various strategies. On the one hand, a persistent dominance of volatility supply tends to suppress expectations for future realised volatility. This dynamic reduces the appeal of trades within the long volatility family, as their expected hit ratio diminishes. It also creates headwinds for Relative Value strategies, given the way they are structured — often with asymmetric payoff profiles that inherently carry a long volatility bias.
On the other hand, low implied volatility opens up opportunities to construct long volatility trades with highly attractive risk profiles. When implied vol is depressed, the cost of options — i.e., the premium paid — is also low. This limits the potential downside of the trade while preserving significant upside if volatility spikes, thereby enhancing asymmetry and improving the overall risk-reward profile.
Throughout 2023, these opposing forces were in constant tension. There were moments when the balance tipped decisively in favor of tactical long volatility positioning. In particular, during the second quarter and again toward the end of the third quarter, we observed conditions where equity volatility appeared extremely underpriced based on our internal assessment. In those instances, the combination of compelling risk profile and favorable risk-reward dynamics justified stepping up our exposure to long volatility trades.
JP Morgan reached a similar conclusion after analysing 100 macroeconomic variables. Their model estimated that the fair value of the VIX, as of December 2023, should be close to 20 — significantly above the actual level of 12.35 at the time of their publication. What is particularly striking is that every single macro variable in their analysis pointed to a higher VIX. [Figure 13]
In short, for most of the year, equity volatility and credit spreads appeared disconnected from macro fundamentals and the prevailing monetary policy regime. One potential explanation for this divergence lies in the unusually delayed transmission of monetary policy. The wealth effect stemming from elevated asset prices, widespread access to ultra-low fixed-rate mortgages, and substantial household and corporate savings have collectively cushioned the economy from the tightening cycle. As a result, the full impact of higher interest rates is likely yet to materialise.
The dynamics of the short volatility trade, which lead to persistently low implied and realised volatility, create a self-reinforcing loop. While initially beneficial for volatility sellers, this mechanism gradually compresses market fluctuations into an increasingly narrow range — ultimately building systemic risk. [Figure 13]
The first layer of risk lies in the potential for a disorderly unwind of outright short volatility positions. But the concern runs deeper. Strategies that adjust leverage based on volatility, such as Risk Parity, CTA, and Volatility Control, tend to increase their exposure when volatility falls. As a result, a sudden spike in volatility can trigger forced, broad-based deleveraging, amplifying market stress.
In the accompanying illustration, the width of the arrows reflects the volume of volatility being supplied to the market — the deeper into this cycle, the greater the supply, and the more fragile the equilibrium becomes.
In the illustration, the distance between each arrow and the centre represents the level of volatility. As the process unfolds, arrows move closer to the centre, indicating progressively lower volatility and a shrinking range of market equilibrium.
This visual captures how the self-reinforcing nature of volatility supply can initially bring greater market stability, but over time leads to a mechanically unsustainable balance. Eventually, the system becomes so compressed that even a modest shock can trigger a sharp and disorderly unwind — transforming stability into sudden instability.
A well-known historical example of this virtuous-turned-vicious cycle occurred in February 2018, when the VIX market effectively imploded. Excessive volatility supply had driven the VIX to record lows, sustaining an exceptionally narrow trading range. In fact, 2017 marked one of the lowest years on record for realised volatility. This calm was ultimately shattered, reinforcing how dangerous a seemingly stable equilibrium can become. [Figure 15]
2023 was a standout year for equity short volatility strategies, which, by our estimates, delivered even stronger returns than in 2017. Using the VIX Futures Inverse Daily Index as a proxy, performance in 2023 outpaced that of 2017, helping to explain why assets under management in these strategies grew significantly despite a deterioration in their risk-reward profile. The recent strength in returns proved compelling enough to attract continued inflows. [Figure 16]
When market complacency sets in, we typically observe a predictable sequence in the compression of volatility risk premia. It begins with the suppression of implied volatility, the first-order risk premium. This is usually followed by the compression of skew and term structure, representing second-order risk premia, and eventually affects convexity, considered the third-order risk premium.
In our view, the compression of equity volatility is now complete, with skew and term structure not far behind. As illustrated in [Figure 17], global equity skews have dropped significantly, now sitting well below their levels from 2020 — a clear sign of how far the repricing of risk has already progressed.
Term structures are also now quite a lot lower than they were in the past at similar levels of implied volatility. This shows that the risk premium is being sucked out of the term structure. [Figures 18 and 19]
Equity convexity hasn’t been attacked yet after the 2020 wipe-out and remains at elevated levels historically. Currency markets, however, paint a different picture where convexity has fully repriced to pre-2020 levels, and there are some attractive entry points to build long convexity positions, notably in EURUSD. [Figure 20]
When examining the long-term macro volatility cycle, we can identify a clear sequence of stages that trace the progression from volatility explosion to compression. These stages vary significantly in length, sometimes lasting just a few months, other times extending over several years. For illustrative purposes, we refer to the period between 2007 and 2017 to map this dynamic.
Stage 0: Explosion
This phase is triggered by a market shock that drives large, disorderly moves and a spike in realised volatility. The catalyst may be fundamental — such as a macroeconomic or geopolitical event — or technical, emerging from a breakdown in local market equilibrium due to excessive leverage, concentrated positioning, or structural fragility. Market participants are caught off guard and rush to buy protection in illiquid conditions, pushing implied volatility sharply higher. The term structure inverts, with front-end volatility becoming more expensive than longer-dated vol, and skew and convexity move to extreme levels.
Stage 1: Normalisation
Markets begin to stabilise, though conditions remain turbulent due to the aftershocks of Stage 0. Liquidity and risk appetite are still impaired. Realised volatility starts to decline, followed by implied volatility, though both remain elevated. The term structure reverts to a more typical upward slope, and volatility risk premia remain rich, continuing to attract interest.
Stage 2: Compression
As high volatility premia draw in sellers, the supply of volatility increases, initially targeting the easiest opportunities. As volatility settles, short vol trades become increasingly profitable, drawing more capital and reinforcing the cycle. Over time, the market exhausts surface-level opportunities, and attention shifts to term structure and skew. System-wide leverage rises, and signs of crowding begin to emerge. Although the market can hold this equilibrium for a time, it becomes increasingly vulnerable to an external shock.
Stage 3: Complacency
The original shock fades into the background, and risk indicators begin flashing green. Both implied and realised volatility reach very low levels, while risk models such as Value at Risk begin excluding the most volatile historical data, making previously unattractive strategies seem appealing again. Markets enter a phase of full-blown complacency, where even convexity is systematically sold. Leverage increases dramatically, crowding becomes extreme, and the system becomes so finely balanced that even a minor disruption can trigger a new volatility explosion — no fundamental shock required.
Based on our empirical and technical analysis, we believe that as of early 2024, the market is in the late stage of Phase 2 in the volatility cycle. Both implied and realised volatility are at historically low levels, and term structure and skew have already undergone significant compression. This suggests that the market is now materially exposed to a fundamental catalyst, though current positioning does not yet appear toxic enough to trigger a self-induced unwind. If this environment persists without an external shock, we are likely to transition into Stage 3 — the complacency phase.
One important takeaway from 2023 is how rapidly the market shifted from Stage 1 to late Stage 2, a transition that unfolded much faster than anticipated during the second quarter. This unexpected acceleration has since been integrated into our analytical framework, allowing us to more accurately track future shifts in the volatility regime.
2B. CREDIT OUTLOOK – CLIMBING WITH LESS CUSHION
In 2023, credit markets demonstrated notable resilience, which we believe is partly due to the fact that they have yet to face a true test. Both households and corporates drew on substantial savings buffers, helping to cushion the immediate impact of tighter monetary policy. This underpins our view that the full effects of rate hikes are likely to materialise with a longer-than-expected lag.
JP Morgan notes that, historically, default cycles tend to peak one to two years after the end of a monetary tightening phase. [Figure 22] However, given current conditions, it's possible that the lag could be even longer this time.
So far, corporates have managed to absorb the impact of higher rates, but the balance between supply and demand may begin to shift. According to JP Morgan, refinancing needs for U.S. high-yield issuers are projected to rise by 28% in 2024 compared to 2023, a development that could start to put upward pressure on credit spreads. [Figure 23]
2C. CROSS-ASSET CORRELATION OUTLOOK – DIVERSIFICATION REMAINS CRITICAL
Diversification is a cornerstone of ADAPT’s portfolio construction philosophy. We strive to achieve a robust and acyclical portfolio by diversifying across multiple dimensions — including asset class, region, product type, strategy, and liquidity profile. This multidimensional approach is essential to building resilience and maintaining performance across different market environments.
In line with our investment mandate and risk management framework, we maintain a careful balance between long volatility/long convexity trades and those that are long status quo/long carry. However, given the dynamics outlined in this report, our analysis in 2023 led us to overweight trades with long volatility profiles, as they offered a more compelling risk-reward in the prevailing market environment.
The insights gained in 2023 have further sharpened our approach going into 2024. We are now better prepared to navigate scenarios involving a much faster collapse in volatility, as well as the extended lag in monetary policy transmission — particularly in how it affects credit spreads and equity volatility.
Shifting perspective from a manager's lens to that of an allocator, ADAPT is designed to deliver uncorrelated returns, serving as a true diversifier within a broader portfolio. We have long advocated for the inclusion of such strategies, especially in an environment where traditional diversification may break down. As the correlation between equities and bonds increases, the classic 60/40 portfolio can become risk additive rather than risk mitigating.
When comparing this traditional allocation to our ‘Next Decade’ portfolio, which replaces half of the bond allocation with ADAPT, the benefits of a genuine diversifying solution become evident. Our objective is to add value through uncorrelated performance and to reduce overall portfolio risk for allocators seeking a more resilient investment framework.
The Next Decade portfolio is outperforming the traditional 60/40 allocation, delivering equity-like returns while maintaining risk metrics — such as maximum monthly loss and maximum drawdown — that are more in line with fixed income. In our view, this clearly demonstrates the value of incorporating a genuinely uncorrelated performance engine into a portfolio, enhancing returns without compromising on risk.
TO CONCLUDE, 2023 was a challenging year for ADAPT’s performance, but one that brought exceptionally valuable lessons. First, we witnessed a compression of risk premia at nearly twice the speed of previous cycles — a pace we had not thought possible, particularly in the context of such a restrictive rate environment. Second, we observed that recent performance momentum increasingly dominates allocation decisions, often taking precedence over traditional risk-reward considerations — as evidenced by the strong preference for short volatility strategies over long fixed income. Third, we came to fully appreciate how immune equity volatility can become to macro and rates-driven volatility, when technical flows are strong enough to overwhelm fundamental signals.
These three unexpected dynamics from 2023 have now been integrated into our framework and will shape our approach going forward.
One conviction remains firmly intact: the cyclical nature of risk premium explosion and compression will continue. This cycle remains a foundational source of opportunity for ADAPT — across short, medium, and long-term horizons. While the character of opportunities evolves over time, our opportunistic, patient, and diversified strategy is designed to adapt and thrive at any point in the cycle.
As always, we welcome your questions and would be happy to provide further insight into the strategy.
January 2024
Updated version as of May 2025