3 ms·

> If you define it as "the probability that a large order can complete", then liquidity has NOT gone up. People seem to have this idea that back in the days of

> If you define it as "the probability that a large order can complete", then liquidity has NOT gone up. People seem to have this idea that back in the days of floor-based trading you could just call up the NYSE and say, "sell 1 million shares of Citigroup!" and the market makers would just kindly oblige you, without widening their spreads or trying to eke out a bigger gain from a transaction which inherently carries significant risk for them. It's possible that the situation for executing giant orders in one fell swoop hasn't gotten any better under HFT, but it hardly seems to have made it worse. And as the large flow traders become more sophisticated and start to implement algos of their own, they'll get better at moving volume at a fairer average price. > With HFT, it's the same 100 shares/futures changing hands thousands of times per day, and they disappear often in times of uncertainty. I don't know about "often," unless you're talking about trading individual issues where news events introduce a high degree of uncertainty. There's really only been one case of a market-wide drying up of liquidity, which lasted for about 15 minutes during the flash crash. There's also nothing new about market-maker liquidity drying up in times of uncertainty or severe volatility. Plenty of stocks went "no bid" during the 1929 crash. Floor-based market makers are no more interested in standing in front of freight trains than algos are. If you're looking for someone/something to blame when markets crash and there are no bids to be found, focus on the Fed and its attempts to manipulate the credit cycle, which periodically fail in spectacular fashion.

tanzam75 13y ago
> It's possible that the situation for executing giant orders in one fell swoop hasn't gotten any better under HFT, but it hardly seems to have made it worse. Plenty of stocks went "no bid" during the 1929 crash. 1929 is an obsolete example. Heck, the SEC didn't even exist in 1929. A more relevant example is 1987 -- how many stocks went "no bid" on Black Monday? Here's how HFTs operate today. Bid furiously while the going is good. If you find yourself on the wrong side of a trade, you can always hit up the actual market-makers for a penny loss. And if the markets collapse, you pull out. What could be better? Supply liquidity when it's not needed, and stay out precisely when liquidity is most needed. Meanwhile, the registered market-makers are pulling back, or pulling out. And why not? They used to count on having the good times to balance out the bad. But now, the HFTs are siphoning off their profits during good times. Thus, the cross-subsidy has gone away. In periods of market stress, the HFTs go hide under a rock and don't participate in absorbing losses. The effect is: There are now fewer market-makers in periods of market stress. That's how HFTs have made things worse.
beagle3 13y ago
> People seem to have this idea that back in the days I don't know who these people are. But many people today see the screens and volumes, and say "see, HFT gives liquidity!", and that's wrong. I'm not saying HFT took liquidity away (it may have, I don't know, but for sure I didn't claim that). But it does not improve the liquidity that matters to most market participants. > executing giant orders in one fell swoop hasn't gotten any better under HFT, but it hardly seems to have made it worse. I agree. I'm just countering the oft repeated meme that "HFT provides liquidity". Essentially, it doesn't. > I don't know about "often," unless you're talking about trading individual issues where news events introduce a high degree of uncertainty. There's really only been one case of a market-wide drying up of liquidity, which lasted for about 15 minutes during the flash crash. There's been tens of "flash crash" and "flash smash" events in the last couple of years. Not of the magnitude of the one blamed on Waddell and Reed, but it's still happening. > There's also nothing new about market-maker liquidity drying up in times of uncertainty or severe volatility. I agree. I just disagree that HFT are providing any kind of service or benefit to the market. They only benefit themselves, at a cost to the market that goes way beyond the profit from arbitrage/scalping action that they do. And it is easy to put a stop to: Adopt the rule Eurex and LIFFE had in 2001 that requires 10:1 order:execution ratio.
ckdhtc 13y ago
"I don't know about "often," unless you're talking about trading individual issues where news events introduce a high degree of uncertainty. There's really only been one case of a market-wide drying up of liquidity, which lasted for about 15 minutes during the flash crash." Actually not true. There was plenty of liquidity during the flash crash. It's just that the people who were selling ended up not liking the price they got. During the flash crash, the market was quite in order, except for a few stocks where all the bids were got (too much sell pressure) and the remaining bids were at $0.01... For most stocks, the market was just fine, just that the spread widened up considerably to take into account the temporary uncertainty in the market. Also, the flash crash took 5 minutes, not 15. Flash crashes happen in human-traded markets took. Recently there was a flash crash in Indian stocks, where humans mark the trades. In fact, human-traded markets are worse: in 1987, the brokers just stopped picking up the phones, even if technically they had an obligation to do so.

tanzam75 13y ago

> It's possible that the situation for executing giant orders in one fell swoop hasn't gotten any better under HFT, but it hardly seems to have made it worse. Plenty of stocks went "no bid" during the 1929 crash. 1929 is an obsolete example. Heck, the SEC didn't even exist in 1929. A more relevant example is 1987 -- how many stocks went "no bid" on Black Monday? Here's how HFTs operate today. Bid furiously while the going is good. If you find yourself on the wrong side of a trade, you can always hit up the actual market-makers for a penny loss. And if the markets collapse, you pull out. What could be better? Supply liquidity when it's not needed, and stay out precisely when liquidity is most needed. Meanwhile, the registered market-makers are pulling back, or pulling out. And why not? They used to count on having the good times to balance out the bad. But now, the HFTs are siphoning off their profits during good times. Thus, the cross-subsidy has gone away. In periods of market stress, the HFTs go hide under a rock and don't participate in absorbing losses. The effect is: There are now fewer market-makers in periods of market stress. That's how HFTs have made things worse.

beagle3 13y ago

> People seem to have this idea that back in the days I don't know who these people are. But many people today see the screens and volumes, and say "see, HFT gives liquidity!", and that's wrong. I'm not saying HFT took liquidity away (it may have, I don't know, but for sure I didn't claim that). But it does not improve the liquidity that matters to most market participants. > executing giant orders in one fell swoop hasn't gotten any better under HFT, but it hardly seems to have made it worse. I agree. I'm just countering the oft repeated meme that "HFT provides liquidity". Essentially, it doesn't. > I don't know about "often," unless you're talking about trading individual issues where news events introduce a high degree of uncertainty. There's really only been one case of a market-wide drying up of liquidity, which lasted for about 15 minutes during the flash crash. There's been tens of "flash crash" and "flash smash" events in the last couple of years. Not of the magnitude of the one blamed on Waddell and Reed, but it's still happening. > There's also nothing new about market-maker liquidity drying up in times of uncertainty or severe volatility. I agree. I just disagree that HFT are providing any kind of service or benefit to the market. They only benefit themselves, at a cost to the market that goes way beyond the profit from arbitrage/scalping action that they do. And it is easy to put a stop to: Adopt the rule Eurex and LIFFE had in 2001 that requires 10:1 order:execution ratio.

ckdhtc 13y ago

"I don't know about "often," unless you're talking about trading individual issues where news events introduce a high degree of uncertainty. There's really only been one case of a market-wide drying up of liquidity, which lasted for about 15 minutes during the flash crash." Actually not true. There was plenty of liquidity during the flash crash. It's just that the people who were selling ended up not liking the price they got. During the flash crash, the market was quite in order, except for a few stocks where all the bids were got (too much sell pressure) and the remaining bids were at $0.01... For most stocks, the market was just fine, just that the spread widened up considerably to take into account the temporary uncertainty in the market. Also, the flash crash took 5 minutes, not 15. Flash crashes happen in human-traded markets took. Recently there was a flash crash in Indian stocks, where humans mark the trades. In fact, human-traded markets are worse: in 1987, the brokers just stopped picking up the phones, even if technically they had an obligation to do so.