It took 32 years, but exchange-traded fund liquidity fear has gone ouroboros, coming full circle to eat its own tail. The Wall Street Journal warned in a recent article that authorized participants are now allowed to hold naked short positions in ETFs, and claims such practices could add to selling pressure in a market crash, and even put them at risk of failing.

But that’s backwards. Any naked short institution is making lots of money in a market crash, absorbing sell orders from others and reducing overall selling pressure. It’s levered long holders that generate counterparty risk and selling pressure in a crash.

The article mixed up two legitimate concerns, which are actually trade-offs. The first is that in a market crash, investors may not be able to trade. Investors could face unaffordable losses, leading to defaults, which could then pass unexpected losses onto others and cause more defaults, turning a market decline into a financial disaster. The second is that in volatile markets, investors may trade at “bad” prices. The extreme is a flash crash in which prices quickly fall and just as quickly rise back to original levels. More commonly there is a net move, but in the volatile trading before the size of that net move is known, investors make trades they regret the next day.

To mitigate the first concern, exchanges invest in elaborate trading mechanisms and computers to ensure orderly order flow and efficient execution even when investors are disorderly and inefficient. For the second, regulators mandate circuit breakers and other features to allow investors to take deep breaths before making unwise trades. Unfortunately, these fixes work at cross-purposes.

ETFs are meant to help on both fronts simultaneously. The Securities and Exchange Commission stimulated the invention of ETFs in reaction to the 1987 equity crash. Markets were overwhelmed in part because there was no good way to trade the entire market at once. Futures failed because they were traded on Chicago futures exchanges while equities were cleared and mainly traded in New York. Data lags, settlement issues and trading hour mismatches exacerbated the crash. Also futures are levered which can add to problems in volatile markets. The SEC wanted a fully collateralized security that traded as an equity.

On Oct. 19, 1987, over 600 million shares traded on the New York Stock Exchange, more than three times normal volume at the time and a record. About 70 per cent of trades were investors trying to adjust overall market exposure, not to buy or sell individual stocks. This includes panicked investors, program traders and futures arbitrage shops. The trading volume overwhelmed systems and kept reported prices out-of-date, triggering more panic and fooling the relatively dumb computer trading systems of the time. People oversold the most liquid names, leading to price mismatches throughout the market.

The SEC wanted to separate market exposure traders from individual stock traders. For one thing, that could replace 500 individual share trades with one S&P 500 ETF trade. For another, it meant the overall market panic wouldn’t hit the most liquid names selectively, nor cause other mismatches in prices among individual stocks. The futures market could stay in sync with the cash market. The hope was simultaneously to promote orderly high-volume execution and to reduce the number of off-market trades, something that other fixes found to be opposing goals.

The first ETF, the SPDR S&P 500 ETF Trust, was launched after five years of negotiations. In the quarter-century since, ETFs have been a positive force by reducing trading costs, improving market efficiency and adding liquidity to the market. So why have critics suddenly decided that the original key design feature of ETFs is a flaw?

Some reasons are rational. The actual record of ETFs in market disruptions of the last five years has been mixed, especially in August 2015 when a volatile market exposed some issues with ETF pricing and trading. This argues that ETFs can be improved, and cautions that they won’t solve all market problems. Also the original conception of a single market factor ETF has evolved into a huge variety of specialized ETFs which might fragment liquidity in a crash rather than concentrating it.

But the main reason is transference. It is entirely rational to fear panics and crashes. They happen. They hurt. ETFs do not prevent them, nor anesthetize the pain. There is good reason to think ETFs can help limit crashes to the fundamental economic realignment that drives them, reducing collateral damage and contagion. But that’s by no means certain. Each crash teaches us new stuff, and sometimes the new stuff is the opposite of what we expected. So people transfer their fear of financial meltdowns and pin it on ETFs. ETFs are designed to be our friends in a crash, and economic theory argues that they will be, but they’re new and not fully tested and bad times sometimes cause us to turn on our friends.