Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Algorithmic trading -- the positive side

physicsoffinance.blogspot.com

Post reply on HN
Print view View original

11–20 of 25 posts

Re: Algorithmic trading -- the positive side

#11

Post by fr0sty »

cma wrote: post #7
markbnine wrote: post #3

Earlier quoted context omitted.

But the paper also rightly ends with a further caveat: While we do control for share price levels and volatility in our empirical work, it remains an open question whether algorithmic trading and algorithmic liquidity supply are equally beneficial in more turbulent or declining markets.

Its pretty much a tax; take out all the day traders and whether you gain or lose due to the spread is more or less random; day traders and hft traders narrow the spread, but since whether you gained or lost was random and they are now capturing a piece of that spread, they (err, ashamedly, we) are basically just taxing everyone. A nominal transaction tax would eliminate most of the ultra-high frequency stuff and coul…

In what way is gaining or losing the spread "random"?

Buyers pay the offer, sellers the ask. If you want to save the spread you can leave a resting order and wait for someone to take you out but you have no guarantee of a fill (or at least a timely one).

Re: Algorithmic trading -- the positive side

#12

Post by copper »

yummyfajitas wrote: post #9
georgieporgie wrote: post #5

My understanding of liquidity in the stock market is that it refers to being able to quickly sell stock, and minimally affect the price of said stock by the sale. I see the value of the former, having an asset you can't sell means its value is rather pointless, but I'm not sure I seed the point of the latter. Isn't money made in the stock market by price volatility? Doesn't algorithmic trading simply smooth out price…

Doesn't algorithmic trading simply smooth out price fluctuations to the point that individual traders receive nothing, while HFT houses skim immense numbers of tiny slivers? Yes. Good speculation smooths out price fluctuations to the point that bad speculators receive nothing, while good speculators receive all the alpha. This is true not only of HFT, but of any good strategy. ...claims of HFT value are heavily under…

> I.e., if an HFT pushes accenture up from $0.05 to $1.00 and sells at $35, following which accenture eventually goes up to $40, they run the risk of having their $1.00 buy trade broken. Then they are stuck with a short sale at $35, while the price of accenture went up to $40.

Shouldn't both trades be broken, though? Breaking just one sounds like the kind of behaviour that would be a strong disincentive to trade at all in the first place.

Re: Algorithmic trading -- the positive side

#13

Post by nandemo »

cma wrote: post #7
markbnine wrote: post #3

Earlier quoted context omitted.

But the paper also rightly ends with a further caveat: While we do control for share price levels and volatility in our empirical work, it remains an open question whether algorithmic trading and algorithmic liquidity supply are equally beneficial in more turbulent or declining markets.

Its pretty much a tax; take out all the day traders and whether you gain or lose due to the spread is more or less random; day traders and hft traders narrow the spread, but since whether you gained or lost was random and they are now capturing a piece of that spread, they (err, ashamedly, we) are basically just taxing everyone. A nominal transaction tax would eliminate most of the ultra-high frequency stuff and coul…

Wait, what?

Individual traders who are value-oriented or technical traders, as well as mutual funds etc, gain money on "buy low, sell high". Of course mutual funds and the like do so on a very long term basis. In any case, they pay the spread. So if the spread is narrower then they make more money in average.

Example: if the Apple's quote is 95/105 (buy/sell) today and 105/115 tomorrow, an invidual trader buying and selling would just break even (buy at 105 and sell at 105) even though Apple's notional fair value has risen. But if there are more market makers, the quotes would be 99/101 and 109/111. Then our individual trader would make 109 - 99 = 10 in the same market conditions.

You're right that a wider spread wouldn't matter much to an individual trader, but that's mainly because their main cost is trading fees (paid to the broker) anyway. For mutual funds, though, a 1 basis-point change in the spread means a lot of money.

Re: Algorithmic trading -- the positive side

#14

Post by yummyfajitas »

copper wrote: post #12
yummyfajitas wrote: post #9

Earlier quoted context omitted.

Doesn't algorithmic trading simply smooth out price fluctuations to the point that individual traders receive nothing, while HFT houses skim immense numbers of tiny slivers? Yes. Good speculation smooths out price fluctuations to the point that bad speculators receive nothing, while good speculators receive all the alpha. This is true not only of HFT, but of any good strategy. ...claims of HFT value are heavily under…

> I.e., if an HFT pushes accenture up from $0.05 to $1.00 and sells at $35, following which accenture eventually goes up to $40, they run the risk of having their $1.00 buy trade broken. Then they are stuck with a short sale at $35, while the price of accenture went up to $40. Shouldn't both trades be broken, though? Breaking just one sounds like the kind of behaviour that would be a strong disincentive to trade at a…

Shouldn't both trades be broken, though?

Under current market rules, no. Besides, this would quickly become a combinatorial disaster. Think of your counterparty who bought at $35 and sold at $36 - now one leg of his trade gets broken, and he has a short position. Or maybe we break both of his trades? Where does the chain end?

And of course, this still hurts people doing stat arb. If you want to go long accenture, short IBM, and your accenture trade is broken, you find yourself with an unhedged IBM short.

Breaking just one sounds like the kind of behaviour that would be a strong disincentive to trade at all in the first place.

This is why HFT's pull out of the market under circumstances where broken trades become likely. Most of the HFT's that stayed in the market during the flash crash made huge money - volatility rocks.

Re: Algorithmic trading -- the positive side

#15

Post by copper »

yummyfajitas wrote: post #14
copper wrote: post #12

Earlier quoted context omitted.

> I.e., if an HFT pushes accenture up from $0.05 to $1.00 and sells at $35, following which accenture eventually goes up to $40, they run the risk of having their $1.00 buy trade broken. Then they are stuck with a short sale at $35, while the price of accenture went up to $40. Shouldn't both trades be broken, though? Breaking just one sounds like the kind of behaviour that would be a strong disincentive to trade at a…

Shouldn't both trades be broken, though? Under current market rules, no. Besides, this would quickly become a combinatorial disaster. Think of your counterparty who bought at $35 and sold at $36 - now one leg of his trade gets broken, and he has a short position. Or maybe we break both of his trades? Where does the chain end? And of course, this still hurts people doing stat arb. If you want to go long accenture, sho…

Thanks, thats interesting. It seems unfair to use the HFT to create stability, and then penalize it for that. I suppose that can't be fixed easily, though.

Re: Algorithmic trading -- the positive side

#16

Post by joshu »

joe_the_user wrote: post #4
gabaix wrote: post #2

"The conclusion, then, is that algorithmic trading (at least in the time period studied, in which stocks were generally rising) does improve market efficiency in the sense of higher liquidity and better price discovery."

Ah, But is it real liquidity or the illusion of liquidity? Liquidity is, more or less, money ready to be invested. The question of whether sophisticated strategies really provide this is complex, like the strategies themselves. If you want background, I think Doug Noland's Credit Bubble Bulletin has done a good job of addressing these questions over the years. At the same time, I think we can see simple things. The b…

No. Liquidity is the measure of the ease with which it is easy to trade a given item. That is to say, shares of IBM are pretty liquid, treasury bonds are very liquid, and dollars are extremely liquid. But your car is not particularly liquid.

The rest of your analysis is nonsense.

Re: Algorithmic trading -- the positive side

#17

Post by cma »

nandemo wrote: post #13
cma wrote: post #7

Earlier quoted context omitted.

Its pretty much a tax; take out all the day traders and whether you gain or lose due to the spread is more or less random; day traders and hft traders narrow the spread, but since whether you gained or lost was random and they are now capturing a piece of that spread, they (err, ashamedly, we) are basically just taxing everyone. A nominal transaction tax would eliminate most of the ultra-high frequency stuff and coul…

Wait, what? Individual traders who are value-oriented or technical traders, as well as mutual funds etc, gain money on "buy low, sell high". Of course mutual funds and the like do so on a very long term basis. In any case, they pay the spread. So if the spread is narrower then they make more money in average. Example: if the Apple's quote is 95/105 (buy/sell) today and 105/115 tomorrow, an invidual trader buying and…

What if the only participants were mutual funds and individual traders with no significant market microstructure strategies. Half of all orders would still be resting orders, the other half would hit the bid/offer in some way. Sometimes you'd lose more due to this, sometimes you'd make more. But it would be a "fair" coin.

Let's say on average it was $0.01 per dollar. As soon as someone entered the market with some microstructure expertise and strategy, everyone but the microstructure guy starts making (say) $0.005 half the time and losing $0.006 the other half.

Reducing some of the randomness and providing more consistency is certainly worth something; you can imagine an example where it was +/- $0.50 per dollar reduced to +$.005, - $.006, but note (assuming the same dollar volume) that the HFT guy would still be making the same amount of money as in the previous example, where he was providing much less value. I.e. the cost that HFT commands isn't strongly tied to the value it provides. This is without even delving into the latency-arbitrage arms race cluster fuck.

Re: Algorithmic trading -- the positive side

#18

Post by antrix »

georgieporgie wrote: post #5

My understanding of liquidity in the stock market is that it refers to being able to quickly sell stock, and minimally affect the price of said stock by the sale. I see the value of the former, having an asset you can't sell means its value is rather pointless, but I'm not sure I seed the point of the latter. Isn't money made in the stock market by price volatility? Doesn't algorithmic trading simply smooth out price…

> minimally affect the price of said stock by the sale. [snip] I'm not sure I seed (sic) the point of the latter

This generally affects sales involving large quantities. Say you want to sell 10000 of IBM. The current order book at the exchange looks like:

    165.55 -  10    
    165.54 -  20    
    165.53 - 200
    165.52 - 500
    165.51 - 100
As you can see, the current market is 165.55 and you decide to place a SELL order for your 10000 IBM. You have two options - place a limit order or a market order.

Market Order: As soon as you place your order, you are going to sweep the order book above. End of those 5 transactions, the market has moved to 165.51 while you've only liquidated 830 of your 10000! Worse, people are noticing a lot of sell activity so the folks who wanted to buy start placing bids at lower & lower prices. By the time you liquidate all your 10000 stocks, the average price you end up getting would be 165 or 164 or lower! You've moved the market by virtue of your sale & made less money as a result.

Limit Order: If you place a limit order for SELL 10000@165.55, then you sell the first 10 to the current highest bid. After that, everyone knows there's a seller looking to offload 10000 units. So they lower their bids. You are worse off!

By now, I hope it is clear why you should care about not moving the market by your sale. The way to solve this problem is to sell in small chunks periodically in ways that does not signal to the market what your actual quantity is. As you can imagine, computers are pretty good at doing this kind of grunt work. Hence Algorithmic trading :-) Specific examples would be VWAP, TWAP, etc. Just Google for them.

PS: this has nothing to do with HFT algos.

Re: Algorithmic trading -- the positive side

#19

Post by jessedhillon »

puzzler314 wrote: post #10
cma wrote: post #7

Earlier quoted context omitted.

Its pretty much a tax; take out all the day traders and whether you gain or lose due to the spread is more or less random; day traders and hft traders narrow the spread, but since whether you gained or lost was random and they are now capturing a piece of that spread, they (err, ashamedly, we) are basically just taxing everyone. A nominal transaction tax would eliminate most of the ultra-high frequency stuff and coul…

I'm not sure a nominal transaction tax would eliminate high frequency trading. HFT guys already pay fees to the exchanges and regulators. Either way, you're right, an increased fee would simply be passed to investors in the form of an increased spread.

Some HFTers execute thousands of trades a day which are cancelled within a few ms -- they are just pinging the market to see how liquidity is at certain price levels, or looking for an opportunity to get out in front of a large quantity trade. This would prevent or impair such traders.
Post reply on HN