The quarterly Goldman Hedge Fund Trend Monitor, aka the HF groupthink update, is released, chock full of HF holding trivia, such as that should Apple ever miss its priced to absolute perfection business model, a whopping 181 hedge funds are going to suffer, and 75 HFs, who have Apple as a top 10 holding, are going to get crushed. Also, we uncover the latest top 10 hedge funds ranked by equity assets (DE Shaw, RenTec and Paulson are the new top 3, although with 2,048 and 2,669 holdings for the first two, they are now receiving 2 and 20 for their quant models which as the NYT highlighted recently no longer work). On the other end of the quant spectrum, are the traditional hedge funds, and as of Q2, the typical fund had an average of 63% of its long-equity assets invested in its 10 largest positions, compared to 30% for a typical large-cap mutual fund, 17% for a small-cap mutual fund, 19% for the S&P and just 2% for the Russell 2000. The top 5 most concentrated hedge fund holdings are AutoNation (46% of market cap held by HFs), Sears (45%), AutoZone (32%), Pactiv (28%) and Novell (27%). Also hilarious perpetual LBO candidate Radioshack has hedge funds make up 24% of its market cap. In other words, any bad news here will kill the stock price faster than a HFT can frontrun the exponential pulling of bids. On the other side, or the names most hated by hedge funds, is Brown Forman, where only 0.2% of HFs make up its market cap, followed by Roper Industries, Stericycle, Hormel, and Praxair. From a surprise upside potential perspective, Goldman estimates that the most HF-shorted names is Crown Media, which has a 99 day short interest ratiom followed by Lifeway Foods, Isramco, K-Fed Bancorp, First South Bancorp, and Costar Group. Shorts Squeezes in these names could be violent. Looking at ETFs, the biggest gross long ETF held by HFs is GLD with $8 billion in long ownership, while the most shorted is SPY with $27.6 billion in shorts, indicating that funds are now "hedging" using this proxy for the entire market. Lastly, in confirmation that hedge funds are for the most part worthless "groupthink" contraptions which merely ride a leveraged beta wave, and suck out management fees, Goldman highlights that the "Most Concentrated" basket of stocks has underperformed the "Least Concentrated" stocks materially since February 2007, confirming that HFs have actually destroyed value in both the past 3 years and YTD, by underperforming the market.
The key findings of the Q2 report update:
Hedge funds’ net weighting in Consumer Discretionary remains the highest of all sectors at 17% followed by Information Technology. Hedge Funds increased their Materials exposure during 2Q 2010 and remained most underweight Consumer Staples and Financials.
Hedge funds have the most positive directional view on Telecom Services given the 65% net long positioning when we combine long and short exposures. Hedge funds appear to hold a 10% net weighting in Materials and are more overweight relative to the Russell 3000 than long exposure alone suggests. Similar portfolio skew occurs in underweights of the Financials and Consumer Staples sectors.
The typical hedge fund allocates 37% of its assets to stocks with less than $2 billion in equity capitalization and 33% to large–caps ($10+ billion). Mid-cap stocks ($2-$10 billion) account for 31% of the average fund’s portfolio. On an aggregate asset basis, hedge funds allocate just 22% of their assets to small-cap stocks, while 43% is allocated to largecap stocks. The difference between the average and the aggregate suggests that the hedge funds with the largest assets under management target large-cap stocks.
Goldman's way of playing the HF hotel game is by creating the "Hedge Fund VIP list", which is characterized as follows:
The hedge fund VIP list (Bloomberg: GSTHHVIP) consists of stocks in which fundamentally-driven hedge funds have a large stake. We define stocks that “matter most” to hedge funds as the positions that appear most frequently among the top ten holdings within hedge fund portfolios. For this analysis, we limit our hedge fund universe to those funds with 10 to 200 distinct equity positions in an attempt to isolate fundamentally-driven investors from quantitative funds or funds that mirror private equity investments.
Our hedge fund VIP list offers investment ideas and tracks long exposure of hedge funds. By construction, the VIP list identifies the 50 stocks whose performance will largely influence the long side of many fundamentally driven hedge funds. The VIP basket lagged the S&P 500 by 1,147 bp during 4Q 2008 (-41% vs. -30%). Since then, the VIP basket performance has reversed, outperforming the S&P 500 by 1,575 bp (41% vs. 26%).
Turnover for the basket during 2Q 2010 was in line with the historical average, with 15 new stocks entering the VIP list compared with a quarterly turnover of 16 stocks since 2001.
From an implementation standpoint, the hedge fund VIP list offers an efficient vehicle for investors seeking to “follow the smart money” based on 13-F filings. The VIP basket has a large-cap bias with a median market capitalization of $39 billion compared with $9 billion for the S&P 500. The VIP list contains stocks from eight of the ten sectors, with Telecommunication Services and Utilities absent. The VIP basket overweights the Information Technology sector (34%) and underweights Consumer Staples (4%).
The basket of the 50 stocks that “matter most” has outperformed the S&P 500 by 71 bp on a quarterly basis since 2001, with a Sharpe ratio of 0.26. As illustrated in Exhibit 3, the VIP list underperformed the S&P 500 during 2Q 2010 by 268 bp (-14.1% vs. -11.4%). Since then, the VIP list has outperformed the S&P 500 by 247 bp (8.8% vs. 6.3%).
Yet possibly the most relevant chart in the Goldman report, is the following which demonstrates that the "Least Concentrated" holdings basket has outperformed the "Most hedge fund Concentrated" stocks massively since February 2007.
We define “concentration” as the share of market capitalization owned in aggregate by hedge funds. The strategy of buying the 20 most concentrated stocks has a strong track record over more than eight years. Since 2001, the strategy has outperformed the market by an average of 312 bp per quarter (not annualized). The back test suggests that this strategy works in an upward trending market but tends to perform poorly during choppy or flat markets. The stocks in the basket tend to be mid-caps (at the lower end of the S&P 500 capitalization distribution), which have outperformed large-caps from 2004 to 2007, contributing to the attractive back-test results. The baskets are not sector neutral versus the S&P 500.
The stocks with the “most concentrated” hedge fund ownership have outperformed the S&P 500 in 2010 ytd by 191 bp (+1.1% vs. -0.8%). The “most concentrated” stocks underperformed steadily for most of 2007 and 2008, but significantly outperformed in 2009. Our “most concentrated” basket outperformed the S&P 500 by 237 bp in 1Q 2010 (+7.7% vs. +5.4%) but lagged by 303 bp in 2Q 2010 (-14.5% vs. -11.4%).
Our “least concentrated” basket has outperformed the S&P 500 in 2010 ytd by 693 bp (+6.1% vs. -0.8%). The “least concentrated” basket outpaced the market by 50 bp in 1Q 2010 (+5.9% vs. +5.4%) and by 440 bp in 2Q (-7.0% vs. -11.4%).
In other words, investors who believe the Fed will ultimately fail in creating an environment in which a simple levered beta strategy will succeed, should short the Most Concentrated stocks, and go long a Least Concentrated basket. These are as follows:
Yet looking at just the long side would provide an incomplete pictures, which is why Goldman also looks at net estimated hedge fund exposure, accounting for short positions.
We combined $593 billion of single-stock and ETF long holdings in 13-F filings of 612 hedge funds with our estimate of hedge fund short positions (based on $434 billion in single-stock, ETF and market index short interest positions filed with exchanges). We estimate hedge funds accounted for 85% of total short interest positions, or $369 billion as of June 30, 2010. Our analysis suggests the typical hedge fund operates 38% net long, down from 43% in 1Q 2010 and 42% at year-end 2009. Part of the short positioning is conducted at the market level via ETFs.
Short positions offer more comprehensive insight to hedge fund sector tilts. Our analysis of short interest data suggests that hedge fund sector net exposure may differ from what 13-F filings indicate.
During prior quarters, short interest data suggested that hedge funds used allocation differences between long and short portfolios to “hedge” their long exposure. So far in 2010, large allocation differences further skew sector positioning relative to the Russell 3000. Hedge fund long exposure shows funds are overweight Materials relative to the Russell 3000 (6.6% versus 4.0%). Short interest data indicates Materials accounts for 4.5% of all short positions. Combining long and short data, hedge funds appear to hold a 10.1% net weighting in Materials and are more overweight relative to the Russell 3000 than long exposure suggests. Similar portfolio skew occurs in underweights of the Financials and Consumer Staples sectors.
Hedge Funds remained underweight Consumer Staples and Financials and increased Materials exposure during 2Q 2010. Hedge funds’ net long weighting in Consumer Discretionary remains the highest of all sectors. Hedge funds appear to underweight Information Technology for the third consecutive quarter (16% vs. 19% in the Russell 3000). Hedge funds were last underweight Information Technology in 2005 although funds were neutral in 2007.
We believe hedge funds account for the vast majority of short positions. The steady growth of shorts in the US equity market during the past eight years has accompanied the rise in hedge fund assets (Exhibits 10 & 11). We estimate that hedge funds account for 85% of all short positions. In the future, mutual funds may become a larger share of the short market, given initiatives such as 130/30 programs. Short interest for the S&P 500 declined over the second half of 2009. Currently, 2.2% of equity cap is held short while the short interest ratio has increased slightly since reaching a 10-year low late last year.
We construct a “typical” long/short hedge fund portfolio. Combining our hedge fund long and short data, we constructed two 50-stock equal-weighted portfolios (one long and one short) in an attempt to replicate a “typical” hedge fund (Exhibits 12 and 13).
We acknowledge certain limitations to our hedge fund short position analysis. There is time delay, as short interest is filed bi-weekly with the exchanges and released with a 10-day delay. The short interest information we have represents positions reported by US broker/dealers. Broker/dealers incorporated outside the US do not have to report their positions. Swaps and other derivatives are also not captured in this analysis.
Using statistical inference and estimates, Goldman has constructed what is a most probable gross long and short portfolio, indicative of prevailing hedge fund groupthink.
Estimated long positions:
Estimated Short positions:
In terms of purchasing trends during Q2, the biggest shift was in holdings of Consumer Discretionary stocks, which declined by 99bps from Q1, yet which still remains the second most overweight sector vs benchmark, next only to Materials. The sector most added to was IT, which saw a 130 bps increase in HF holdings during Q2.The most despised sectors by Hedge Funds continue to be Consumer Staples, Financials and Utilities.
Allocation to Materials rose by 10 bp to 6.9% in 2Q and Materials is now the most overweight sector relative to the Russell 3000 (7% vs. 4%). During 2Q 2010 hedge funds reduced their long exposure to Consumer Discretionary by 99 bp to 14.1%. Exposure to Consumer Discretionary, previously the most overweight sector, fell by 99 bp. Hedge funds are currently 287 bp overweight Consumer Discretionary and 641 bp underweight Consumer Staples relative to sector weights in the Russell 3000. We recognize hedge funds do not manage assets based on a benchmark, but understanding and interpreting sector tilts requires an orientation to some index.
Information Technology and Telecom Services represented the largest increases in hedge funds’ long asset allocation during 2Q (130 bp and 49 bp respectively). Hedge funds decreased exposure to Consumer Discretionary and Financials. Utilities and Telecom Services outperformed the market in 2Q, accounting for some of the higher weighting in portfolios. However, Materials and Info Tech underperformed during the quarter suggesting hedge funds bought shares.
Changes in mutual fund sector allocation were most differentiated from hedge funds in Consumer Staples, Health Care, and Materials. Mutual funds increased exposure to Consumer Staples and Health Care during the quarter and lowered exposure to Materials. Hedge funds decreased exposure to Consumer Staples and Health Care, and increased long asset allocation to Materials.
Sub-sectors: What hedge funds are buying and selling
We estimate the dollars invested by hedge funds in various sub-sectors based on change in ownership during the quarter. Sub-sectors in which hedge funds added to positions slightly outperformed sub-sectors in which they sold during 2Q (-14% vs. -15%). Funds added to positions in Industrials sub-sectors such as Building Products, Security & Alarm Services, and Trucking. The sub-sector with highest outflows was also in Industrials, Heavy Electrical Equipment (Exhibit 15).
It also appears that size is starting to matter less to hedge funds, with 43% of assets invested in stocks with stocks with market cap higher than $10 billion, down from 47% as of Q1, yet in line with Q2 2009 exposure. Based on continuing decreases in stock liquidity, Goldman estiamtes that it expects "the largest hedge funds with multiple billions of dollars in assets under management will increasingly seek investment opportunities among mid- and large-cap stocks given the difficulty of investing large amounts of capital in small-cap companies. This trend may contribute to large-cap outperformance in coming quarters."
Another interesting observation by Goldman: "small funds favor small caps, large funds go for the megas."
The typical hedge fund allocates 37% of its assets to stocks with less than $2 billion in equity capitalization and 33% to large cap stocks ($10+ billion). Mid-cap stocks ($2-$10 billion) account for 31% of the average fund’s portfolio (Exhibit 17).
On an aggregate asset basis, hedge funds allocate just 22% of their assets to smallcap stocks, while 43% is allocated to large-caps. The difference between the average and the aggregate suggests that the hedge funds with the largest assets under management target large-cap stocks.
For comparison, mutual funds invest an average of 13% of their portfolios in smallcap stocks and 59% in large-cap stocks. The mutual fund tilt is even more pronounced on an aggregate asset basis where 10% of assets are invested in small-cap stocks and 65% is in large-cap stocks.
Also, when it comes to the most popular product these days, i.e. ETFs, it appears that hedge funds do not use ETFs for investing purposes at all, but merely to hedge existing exposure, and that primarily as a short hedge: Goldman estimates that the greatest gross exposure is in a short SPY position to the tune of $27.6 billion.
Hedge funds appear to use ETFs more as a hedging tool than as a directional investment vehicle, based on our analysis of 13-F and short interest filings. We estimate that hedge funds hold $101 billion in gross exposure to ETFs compared with $900 billion of gross exposure to single-stocks.
The $81 billion of short ETF positions accounts for 80% of the hedge fund gross ETF exposure. In contrast, single-stock short positions ($307 billion) represent 34% of hedge fund gross single-stock positions. The most shorted ETFs tend to be index hedges (representing $44 billion of the $81 short positions). Commodity-related ETFs appear to be the only ETFs that hedge funds utilize on the long side.
ETFs now represent 3% of long assets, down from 6% in 1Q 2009. This is consistent with a falling correlation environment, in which stock-picking comes into focus.
Yeah, we are also not sure where Goldman imagined that last bit from.
Here is the ETF data in tabular form:
The key summary appendix tables are the following:
Stocks with most number of hedge fund investors:
Stocks with biggest popularity changes in Q2 2010
Stocks with biggest HF concentration increases and decreases:
Biggest hedge fund hotels by popularity band:
The squeeze hunters will want to pay particular attention to this one: it shows the stocks with the highest short interest in the under and over $1 billion mkt cap segments:
Another key observation: hedge fund holdings are extremely concentrated, and a few key moves can make or break a fund:
Hedge fund returns are highly dependent on the performance of a few key stocks. The typical hedge fund has an average of 63% of its long-equity assets invested in its 10 largest positions compared with 30% for the typical large-cap mutual fund, 17% for a small-cap mutual fund, 19% for the S&P 500 and just 2% for the Russell 2000 Index.
And last, the chart many who are part of the HF pissing contest have been waiting for: the 100 largest hedge funds ranked by equity assets: