Applying a robust risk model to hedge fund holdings data helps avoid losses and yields profitable opportunities. In this article, we highlight the sectors with the largest Hedge Fund losses due to crowding in Q3 2015, which sum to $4 billion. Our methodology provides an early-warning system for losses in crowded names. This analysis also identifies crowded stocks beaten up by hedge fund liquidations, which tend to mean-revert.
Analyzing Hedge Fund Sector Crowding
Our edge comes from a central thesis: the most crowded stocks are those that contribute the most to hedge fund stock-specific volatility (volatility of alpha). Furthermore, the direction of this alpha (positive or negative) is a leading indicator. A robust analysis of the AlphaBetaWorks Statistical Equity Risk Model allows us to identify stocks that are the highest contributors to stock-specific volatility for hedge funds in each sector. These are the most crowded stocks that stand to benefit the most from accumulation and stand to lose the most from liquidation.
While a static crowding analysis using our risk model provides valuable insights, we go further by identifying Hedge Fund Aggregate Sector Alpha – the alpha (stock-specific performance) of aggregated hedge fund portfolios by sector. This makes the analysis dynamic: If Hedge Fund Aggregate Sector Alpha is trending up, capital is flowing into crowded stocks. Conversely, if it is trending down, capital is flowing out of crowded stocks – often abruptly. Yes, crowding is good at some times and bad at others. Further, Hedge Fund Aggregate Sector Alpha trends persist for months and years, providing advanced notice of losses. Importantly, crowded stocks hit hard by liquidations tend to mean-revert: the worst risk-adjusted performers often become attractive long opportunities.
Hedge Fund Sector Aggregates
We create aggregate portfolios of hedge fund positions in each sector. Each such sector portfolio is a Hedge Fund Sector Aggregate within which we identify the highest contributors to security-specific (residual) volatility (the most crowded stocks). This follows the approach of our earlier articles on hedge fund crowding.
The Hedge Fund Sector Aggregate Alpha (αReturn, residual, or security-specific return) measures hedge fund security selection performance in a sector. It is the return HF Sector Aggregate would have generated if markets had been flat. αReturn can indicate accumulations and liquidations.
The AlphaBetaWorks Statistical Equity Risk Model, a proven tool for forecasting portfolio risk and performance, estimated factor exposures and residuals. Without an effective risk model, simplistic crowding analyses ignore the systematic and idiosyncratic exposures of positions and typically merely identify companies with the largest market capitalizations.
Sectors with the Largest Losses from Hedge Fund Crowding
During Q3 2015, hedge funds lost $4 billion to security selection in the five sectors below. Said another way: if hedge funds had simply invested passively with the same risk, their sector long equity portfolios would have made $4 billion more. The monthly losses are listed (in $millions) below:
7/31/2015 | 8/31/2015 | 9/30/2015 | Total | |
Other Consumer Services | -101.16 | -113.93 | -312.84 | -426.77 |
Oil and Gas Pipelines | 472.21 | -465.63 | -10.29 | -475.93 |
Specialty Chemicals | -155.87 | 196.41 | -730.73 | -534.32 |
Oil Refining and Marketing | 262.69 | -167.15 | -388.52 | -555.67 |
Semiconductors | -240.71 | -1,422.70 | -660.95 | -2,083.65 |
The Semiconductor Sector was particularly painful for hedge funds in Q3 2015, which we examined in a previous article.
Below we provide our data on three of the above sectors: historical Hedge Fund Sector Alpha and the most crowded names.
Specialty Chemicals – Hedge Fund Alpha and Crowding
Hedge Fund Specialty Chemicals Security Selection Performance
Hedge Fund Specialty Chemicals Crowding
The following table contains detailed data on these crowded holdings:
Exposure (%) | Net Exposure | Share of Risk (%) | |||||
HF Sector Aggr. | Sector Aggr. | % | $mil | Days of Trading | |||
GB:PAH | Platform Specialty Products Corp. | 17.59 | 2.52 | 15.07 | 1,351.8 | 14.3 | 44.62 |
APD | Air Products and Chemicals, Inc. | 47.46 | 13.89 | 33.57 | 3,010.8 | 13.7 | 22.09 |
LYB | LyondellBasell Industries NV | 3.36 | 23.03 | -19.67 | -1,764.2 | -5.9 | 14.04 |
GRBK | Green Brick Partners, Inc. | 2.99 | 0.25 | 2.74 | 245.7 | 79.7 | 10.58 |
GRA | W. R. Grace \& Co. | 11.76 | 3.45 | 8.32 | 745.8 | 11.0 | 2.99 |
PX | Praxair, Inc. | 0.31 | 16.29 | -15.98 | -1,433.5 | -5.9 | 2.21 |
AXLL | Axiall Corporation | 2.79 | 1.20 | 1.59 | 142.8 | 4.5 | 0.74 |
TROX | Tronox Ltd. | 1.80 | 0.45 | 1.35 | 121.2 | 14.2 | 0.36 |
ARG | Airgas, Inc. | 0.19 | 3.77 | -3.59 | -321.8 | -4.1 | 0.33 |
SIAL | Sigma-Aldrich Corporation | 3.32 | 7.88 | -4.56 | -408.6 | -2.3 | 0.28 |
NEU | NewMarket Corporation | 0.23 | 2.61 | -2.38 | -213.4 | -6.0 | 0.26 |
VHI | Valhi, Inc. | 0.02 | 0.91 | -0.88 | -79.2 | -240.2 | 0.26 |
CYT | Cytec Industries Inc. | 0.07 | 2.04 | -1.97 | -176.5 | -2.0 | 0.18 |
ASH | Ashland Inc. | 1.66 | 3.89 | -2.23 | -200.0 | -2.4 | 0.18 |
POL | PolyOne Corporation | 0.19 | 1.65 | -1.46 | -131.2 | -4.3 | 0.10 |
TANH | Tantech Holdings Ltd. | 0.00 | 0.19 | -0.19 | -17.3 | -2.7 | 0.09 |
BCPC | Balchem Corporation | 0.00 | 0.82 | -0.82 | -73.4 | -8.8 | 0.07 |
CBM | Cambrex Corporation | 0.06 | 0.65 | -0.59 | -53.2 | -2.1 | 0.06 |
CMP | Compass Minerals International, Inc. | 0.15 | 1.31 | -1.16 | -104.0 | -4.8 | 0.06 |
… | Other Positions | 0.29 | 0.51 | ||||
Total | 100.00 |
Oil Refining and Marketing – Hedge Fund Alpha and Crowding
Hedge Fund Oil Refining and Marketing Security Selection Performance
Hedge Fund Oil Refining and Marketing Crowding
The following table contains detailed data on these crowded holdings:
Exposure (%) | Net Exposure | Share of Risk (%) | |||||
HF Sector Aggr. | Sector Aggr. | % | $mil | Days of Trading | |||
MWE | MarkWest Energy Partners, L.P. | 18.23 | 5.31 | 12.92 | 848.9 | 6.1 | 31.86 |
VLO | Valero Energy Corporation | 0.38 | 16.06 | -15.68 | -1,030.4 | -2.7 | 23.34 |
TSO | Tesoro Corporation | 14.32 | 5.36 | 8.96 | 589.0 | 1.4 | 12.74 |
TRGP | Targa Resources Corp. | 8.99 | 2.52 | 6.47 | 425.3 | 8.7 | 7.76 |
PSX | Phillips 66 | 9.21 | 21.86 | -12.66 | -831.8 | -2.8 | 6.03 |
PBF | PBF Energy, Inc. Class A | 6.80 | 1.23 | 5.56 | 365.6 | 7.8 | 5.84 |
NGLS | Targa Resources Partners LP | 8.74 | 3.52 | 5.21 | 342.7 | 6.2 | 2.84 |
WGP | Western Gas Equity Partners LP | 3.58 | 6.63 | -3.05 | -200.5 | -7.4 | 2.06 |
MPC | Marathon Petroleum Corporation | 9.59 | 14.34 | -4.75 | -312.0 | -1.1 | 1.81 |
TLLP | Tesoro Logistics LP | 5.12 | 2.33 | 2.79 | 183.1 | 3.5 | 1.45 |
HFC | HollyFrontier Corporation | 1.29 | 4.22 | -2.93 | -192.3 | -1.4 | 1.11 |
WNR | Western Refining, Inc. | 0.21 | 2.10 | -1.89 | -124.5 | -1.4 | 0.61 |
IOC | Interoil Corporation | 0.66 | 1.50 | -0.84 | -55.3 | -6.9 | 0.49 |
GEL | Genesis Energy, L.P. | 4.35 | 2.20 | 2.15 | 141.1 | 6.2 | 0.34 |
ENBL | Enable Midstream Partners LP | 0.39 | 1.73 | -1.34 | -88.2 | -31.6 | 0.33 |
EMES | Emerge Energy Services LP | 0.01 | 0.43 | -0.42 | -27.6 | -6.1 | 0.29 |
DK | Delek US Holdings, Inc. | 0.00 | 1.07 | -1.07 | -70.0 | -1.2 | 0.26 |
WNRL | Western Refining Logistics, LP | 1.57 | 0.36 | 1.21 | 79.5 | 15.0 | 0.24 |
ALJ | Alon USA Energy, Inc. | 0.00 | 0.67 | -0.67 | -44.1 | -2.3 | 0.18 |
NS | NuStar Energy L.P. | 3.50 | 2.33 | 1.17 | 76.9 | 1.4 | 0.15 |
… | Other Positions | 0.07 | 0.28 | ||||
Total |
Semiconductors – Hedge Fund Alpha and Crowding
Hedge Fund Semiconductor Security Selection Performance
Given the magnitude of recent semiconductor sector liquidations and the record of mean-reversions, the following crowded hedge fund semiconductor bets may now be especially attractive:
Hedge Fund Semiconductor Crowding
The following table contains detailed data on these crowded holdings:
Exposure (%) | Net Exposure | Share of Risk (%) | |||||
HF Sector Aggr. | Sector Aggr. | % | $mil | Days of Trading | |||
SUNE | SunEdison, Inc. | 33.18 | 1.82 | 31.36 | 2,550.9 | 9.6 | 86.72 |
MU | Micron Technology, Inc. | 18.87 | 3.95 | 14.93 | 1,214.1 | 2.9 | 8.85 |
INTC | Intel Corporation | 3.72 | 27.94 | -24.22 | -1,970.2 | -1.6 | 2.01 |
SEMI | SunEdison Semiconductor, Inc. | 3.22 | 0.14 | 3.08 | 250.7 | 52.5 | 0.38 |
SWKS | Skyworks Solutions, Inc. | 0.04 | 3.85 | -3.82 | -310.4 | -0.9 | 0.38 |
TXN | Texas Instruments Incorporated | 0.09 | 10.38 | -10.28 | -836.6 | -1.9 | 0.32 |
NXPI | NXP Semiconductors NV | 7.90 | 4.41 | 3.49 | 283.6 | 1.0 | 0.29 |
AVGO | Avago Technologies Limited | 3.29 | 6.69 | -3.40 | -276.3 | -0.5 | 0.18 |
FSL | Freescale Semiconductor Inc | 0.02 | 2.40 | -2.38 | -193.5 | -5.2 | 0.17 |
ON | ON Semiconductor Corporation | 3.39 | 0.97 | 2.42 | 196.6 | 4.3 | 0.08 |
MLNX | Mellanox Technologies, Ltd. | 1.89 | 0.43 | 1.45 | 118.3 | 0.7 | 0.08 |
BRCM | Broadcom Corporation Class A | 7.81 | 5.51 | 2.30 | 187.2 | 0.5 | 0.07 |
MX | MagnaChip Semiconductor Corporation | 0.92 | 0.05 | 0.87 | 70.9 | 31.2 | 0.07 |
ADI | Analog Devices, Inc. | 0.05 | 3.90 | -3.85 | -312.9 | -1.7 | 0.06 |
QRVO | Qorvo, Inc. | 1.13 | 2.32 | -1.19 | -96.7 | -1.1 | 0.06 |
NVDA | NVIDIA Corporation | 0.58 | 2.10 | -1.51 | -123.1 | -0.4 | 0.04 |
GB:0Q19 | CEVA, Inc. | 1.25 | 0.08 | 1.17 | 95.5 | 30.7 | 0.04 |
MRVL | Marvell Technology Group Ltd. | 0.04 | 1.32 | -1.28 | -104.4 | -0.9 | 0.03 |
MXIM | Maxim Integrated Products, Inc. | 0.34 | 1.90 | -1.56 | -126.9 | -1.7 | 0.02 |
MXL | MaxLinear, Inc. Class A | 0.74 | 0.12 | 0.62 | 50.6 | 2.8 | 0.02 |
… | Other Positions | 0.36 | 0.13 | ||||
Total |
Conclusions
- Data on the crowded names and their alpha can reduce losses and provide profitable investment opportunities.
- A robust and predictive equity risk model is necessary to accurately identify hedge fund crowding.
- Allocators aware of crowding can gain new insights into portfolio risk, manager skill, and fund differentiation.
- Crowded bets tend to mean-revert following liquidation: the worst risk-adjusted performers in a sector become the best.