Last month, the Kauffman Foundation published an 85-page report that reached troubling conclusions about exchange-traded funds (ETFs). According to its authors, these popular investment vehicles prevent startups from easily tapping the capital markets, while increasing systemic risk.
ETFs have enjoyed spectacular success. According to the report, in 1995 there were four ETFs with $2.3 billion in market capitalization. At the end of September 2010, there were 3,257 exchange-traded products with a combined market cap of $1.2 trillion. In such a massive market, if there is a problem, it's a big one.
Do ETFs discourage IPOs?
The authors charge that ETFs are "undermining the traditional price discovery role of exchanges," especially for small-cap stocks.
According to the report, the popularity of index funds, and the massive piles of capital they now contain, has made stock prices highly correlated with one another. Because so many ETF investors own a piece of the same basket of stocks, if one stock rises or falls, many others are prone to follow it:
The July 2010 report by Empirical Research Partners, for example, illustrates that the stock prices of large capitalization common stocks now move in the same percentages and in the same direction 60% of the time -- a level exceeded only twice in 84 years (1929 and 2009). ... [C]orrelations have been rising for most of the decade, both in the broad market and within sectors like financials, capital goods, and transportation.
When wildly different stocks tend to move in unison, the market's clearly doing a lousy job of sorting those shares' real value according to their potential risks and returns. This creates a bad environment in which to launch an IPO, because if nearly every stock moves up and down in tandem with nearly every other stock, high-growth startups have a far smaller chance to soar in price when they first go public. Since these tiny, rookie companies are still riskier than big-name blue chips, losing the opportunity for outsized stock price appreciation makes them a lot less attractive to investors. Indeed, fewer IPOs are launching these days.
But is an overabundance of ETFs really causing this unusual correlation? Or could something else be herding the market en masse -- say, the perception of increased systemic risk? After all, correlation levels peaked in 1929, decades before ETFs ever existed. But the Kauffman report's authors point out that we now have an unprecedented mix of high correlation with low volatility. Since the market's not wildly swinging back and forth, they argue, investors' fear of systemic risk can't be making stocks move in lockstep.
This is an interesting theory. On the face of it, the idea that more money in index funds could lead to higher correlation seems plausible, but it's far from proven. Who says you can't have fear of systemic risk without high volatility? Who says that, if you have lots of ETFs and high stock correlations at the same time, ETFs are necessarily the cause, and high stock correlations the effect? Why couldn't something else -- say, the Fed's attempts to reinflate asset bubbles via quantitative easing -- be causing that high correlation instead?
The report suggests that we should allow small-cap companies to refuse inclusion in indices, and therefore in the ETFs and mutual funds which track them, in order to escape the cohesive pricing effect. But if you're the CEO of a company that just went public, wouldn't you want index-tracking funds to feel obliged to buy your shares?
Do ETFs increase systemic risk?
According to the report, under conditions of high volatility, ETFs are vulnerable to collapse:
As more ETFs are created, the risk grows that, in the event of a future market meltdown triggered by any number of possible causes, the rush to unwind the ETFs will aggravate any sell-off. Indeed, some creators of ETFs may not be able to honor their obligations. If those institutions or holders of ETFs are deemed sufficiently important or interdependent with other financial actors, the U.S. government could be forced again to make the agonizing decision whether to come to the rescue, as it did with AIG and a number of other large enterprises during the financial crisis of 2008.
The report argues that a sudden surge in the market could mean that the price of all the stocks that make up an ETF will rise faster than the price of the ETF's own shares, which implies that creating new ETF shares to meet demand would be a losing proposition. And when prices plunge, and investors want to pull money out of an ETF, the report fears that the folks running the ETF won't be able to sell those holdings at a high enough price to cover the cost of the redeemed ETF shares.
The report spends a lot of ink persuasively demonstrating that for many of the stocks held by the iShares Russell 2000 Index Fund (NYSE: IWM), it would take a long time -- as much as 188 trading days, in the worst case -- to buy or sell enough shares to create or retire a 50,000-share unit of the ETF without causing big swings in the price of the underlying holdings.
But while this problem sounds scary in theory, given the way ETFs create and redeem shares in practice, the effects will likely be less than earthshaking.
ETF companies work with Authorized Participants (APs), which handle the share creation/redemption process and act as market makers for the ETF. Suppose one of those APs perceives sufficient market demand for another block of index-tracking shares. First, it assembles the underlying securities for delivery to the ETF company. Only after those holdings are in place will the company issue a block of new ETF shares to the AP. Similarly, if customers are selling more shares than they're buying, the AP can return blocks of ETF shares to the ETF company, in exchange for the underlying securities.
Under normal market conditions, this system self-corrects any price anomalies caused by supply/demand imbalances. If demand for an ETF is high -- if the price per ETF share is consistently higher than the value of the underlying holdings -- the AP has incentive to create additional shares by purchasing enough of the underlying equities, then tendering them to the ETF company for an additional unit. If demand is low, and the price per share lags the underlying value of the equities, the AP has incentive to buy a block of ETF shares, then redeem it with the ETF company in exchange for the underlying equities.
The Kaufman report is right about the APs' potential difficulty in creating or destroying ETF units when the market is highly volatile. But while the APs normally act as market makers for their ETFs, buying and selling surplus shares to ensure smooth trading, they are not obliged to make sure the ETF share value exactly matches that of the underlying securities at all times. If there's a sell-off, and the market irrationally prices the value of the ETF below that of the shares, the AP will normally buy ETF shares. But if the selling becomes a tsunami, and the AP loses confidence that it can sell the underlying securities for enough to cover what it'd pay for blocks of ETF shares, it will temporarily step aside, just as stock market-makers did during the so-called flash crash.
So when the market gets extremely volatile, index ETF shares might gyrate wildly -- just like the stock of any public company. But as long as the company issuing the ETF hangs onto the shares it gets from its APs, so that it can trade them back to the APs as blocks of ETF shares get redeemed, there's no material, inherent systemic risk involved. In short, nobody's going to go bust and need a government bailout.
Did ETF shorting cause the flash crash?
The report cites legitimately scary levels of short interest for some ETFs, including "six ETFs with more than 100% short interest and at least 13 ETFs with more than 5%."
There are two reasons why high level of short interest can be a problem. First, it increases the risk of volatility. When unwarranted selling starts squeezing the price of a stock or ETF, shareholders might panic and start a stampede. And when unexpected good news catches short-sellers off-guard, they may all simultaneously rush to buy shares and cover their positions.
Second, high short interest is intrinsically suspicious. Generally speaking, you can only borrow and short shares that you've bought on margin. The proportion of shares bought on margin (which include fully paid-for shares held in margin accounts) varies widely from equity to equity. However, it's not unusual to have difficulty finding marginable shares to short, even for stocks with less than a 5% short interest (which is considered high).
If short interest gets unusually intense, it may signal that people are selling shares they don't own and haven't borrowed from someone else -- a generally illegal practice known as "naked shorting." Mathematically, when short interest exceeds 100% of outstanding shares, there must be at least some naked shorting going on.
Thus, it's likely that illegal naked shorting's involved in most or all of the 13 high-short-interest ETFs the Kauffman report cites. However, naked short-selling is a problem for plenty of other investments besides ETFs. And in any event, finding short-selling issues with 13 out of 3,257 ETFs hardly constitutes a crisis.
Finally, the authors diligently try to prove that iShares' Russell 2000 Index ETF, along with the S&P 500-tracking SPDR Trust (NYSE: SPY), set off the May 6 flash crash. They convincingly demonstrate that both ETFs were violently affected by the flash crash, along with hundreds of other equities. But they don't seem to have persuasive proof regarding who was in the driver's seat.
Dubious diagnosis, decent prescriptions
If ETFs were more prone than stocks to fail amid high volatility, that would be a major issue. Unless you believe that the report proves that assertion, it appears that most unleveraged long ETFs are no more vulnerable to a black-swan event than any stock in the S&P 500. In fact, since index ETFs spread risk among multiple equities, they're probably less risky than most individual stocks.
However, even though I think the report's criticisms of ETFs are off-base, I do consider some of the proposed remedies properly Foolish. For example, the authors attack the common broker practice of lending out marginable securities from their customer accounts to short-sellers. Suppose you buy shares of SPY, confident that the S&P 500 will rise over the long run. Your broker then quietly lends some of your shares to a short-seller, collecting fees and interest in the process. The broker keeps all the profit from those fees for itself, even though it's allowing the short-seller to use your own shares to effectively bet against the investment you've made.
While you can usually opt out of letting brokers loan your shares to shorts on the sly, most brokers won't advise you to do so. The report suggests that brokers shouldn't be allowed to lend out marginable securities from your account unless you explicitly approve such a move. If you do, the brokers should be obligated to split with you any money they make off the short-seller.
The report also calls for much more stringent enforcement of the rules against failing to deliver shares by the settlement date. (The authors claim that failures to deliver inordinately afflict ETFs; even if they don't, they're still a serious problem in general.) And they call for restrictions on anonymous, off-exchange "dark pool" trading, to eliminate front running and other abuses, and ensure that customer orders get priority.
Plus, they call for the virtual elimination of market orders; every order would be a limit order, either specified by the customer or determined automatically by standard rules. As they see it, market orders increase the potential for, and the likely damage from, flash-crash-like volatility.
The authors are not big fans of Sarbanes-Oxley, which is holy writ for some Fools, but some of their suggested tweaks to the legislation seem worth considering. For example, they propose giving the shareholders of micro-cap companies greater leeway in deciding how much of Sarbanes-Oxley to implement. For some companies, that choice may make access to the capital markets a viable alternative to selling out to a bigger, less entrepreneurial corporation. While some bad actors may slip through, greater flexibility for micro-caps could also enhance the capability of the system to nurture and reward innovation.
In general, it's probably fair to say that with the advent of ETFs, investors are paying more attention to macro-level considerations, and less to individual stocks. As advocates for entrepreneurs, the authors of the Kauffman report are not happy about that. But with America suffering from financial illiteracy, the popularity of ETFs could help more people start considering how big-picture issues such as the decline of the dollar will affect their portfolios.
Since ETFs can help investors form strategies based on the broader movements of the entire economy, it's ironic that these investments should be criticized for creating unintended, far-reaching effects that could potentially frustrate or sabotage those same strategies.
This particular report appears to overestimate the dangers of ETFs. But given these investments' relative newness, and the reality that not all of them have worked as well as investors might have expected, I'm glad to see thoughtful critics raising these sorts of issues.
We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors.
Guest contributor Brad Hessel currently has no position in any of the equities mentioned; however, Brad's clients may have such positions. The Fool's disclosure policy includes certain trading restrictions that apply to Brad. However, his clients are not subject to our disclosure policy, and thus are free to trade any such equities.
