It will be interesting to see if this market provides better pricing for non-HFT participants.
The HFT outfits claim they net out to better pricing for all ("liquidity benefits"), but that's somewhat hard to swallow given that they're acting as giant money sinks on the market system.
HFTs demonstrably are liquidity providers. That's a technical term with a real meaning: liquidity is the ability to trade when you want to trade in the quantity you want to trade it, and it most certainly is not a natural property of the market; in order to buy an instrument, someone has to be willing to sell it.
Meanwhile, what is a "giant money sink", and how is that what HFTs are? I see how HFTs cut out the middlemen who used to profit from volatility, but the low-tech traders they replaced were not themselves value investors.
And formal market makers accept an obligation to stay in the market and provide full-time liquidity in return for their profits. I have to question the value of displacing and de-funding them in favor of HFTs who can walk away at the worst moment (as when they exacerbated the flash crash).
If there are n people in the market moving money around in a closed system then the combined wealth of those n people is constant.
If another person joins in and is making a net profit then it must be the case that the combined wealth of the original n is decreasing.
Admittedly this relies on money not being created or destroyed, which may cause the model to be a poor approximation of reality.
One current big controversy among economists is whether the HFTs themselves have an effect on volatility (there are studies claiming all of "increases", "decreases", and "have little effect", but the evidence for any of those conclusions is weak, and it depends on exactly how you define volatility). A liquidity provider who provided added liquidity on very short time scales, but at the expense of increased volatility at longer time scales, wouldn't necessarily be advantageous for longer-term market participants. That'd be essentially trading one kind of noise for another one--- less millisecond variation, and more occasional giant aberrations, if indeed things like the May '10 "flash crash" were caused by HFTs.
It will be interesting to see if this market provides better pricing for non-HFT participants.
Given that the only difference between this and a normal market is that a bunch of people offering to buy and/or sell at prices more favorable than the current inside bid/ask are banned from the market, I'm going to go out on a limb and say that no, it won't help pricing very much.
Or, rather, what it will do is move profits that would have gone to the HFT firms and shift them to people placing normal, non-speculative/exploitative/high-frequency limit orders.
Which means that really, there will be an arms race to figure out how to place orders that are as exploitative as possible without tripping the banhammer. The people that push the closest that limit will make the most money, but spreads will still likely be quite a bit higher than in an unfettered market if the rules have any teeth at all - long term value traders are not good at keeping spreads thin, I'd guess that in the current markets their effects on spreads are not even measurable at all, they are so insignificant...
Given that the only difference between this and a normal
market is that a bunch of people offering to buy and/or
sell at prices more favorable than the current inside
bid/ask are banned from the market, I'm going to go out
on a limb and say that no, it won't help pricing very
much.
Since trading on Light Pool will be more expensive and
slower for those firms, they’re not likely to use the
ECN, reducing the negative selection investors
experience, Galinov said.
Contributors, which will include long-term investors,
will receive a “significant” rebate when they trade
against orders resting in Light Pool, while neutral
firms, or those whose behavior falls between the other
groups, may or may not get one, Galinov said.
http://ai.eecs.umich.edu/people/wellman/?p=40
Exchange A: Price you can BUY ACME Co. = $34.50
Exchange B: Price you can BUY ACME Co. = $34.45
Exchange B: Price you can SELL ACME Co. = $34.40
Exchange A: Price you can SELL ACME Co. = $34.35Is this possible because the market maker ensures some kind of buffer from the HFT shops in exchange for a bigger spread?
I'm not in this field so I ask from ignorance.
I believe the mechanism going on here is this:
Big Traders offer to buy at $10.00 on Light Pool. Their offer sits there, and gets filled slowly over time. Then, for whatever reason the price moves, and a bunch of speed traders try to sell on INET/ARCA/BATS at $9.99 (perhaps in anticipation of the market moving down to $9.90). Due to the high fees, they don't place those orders on Light Pool. The market crosses for a little while, no trades occur, and Big Traders get the opportunity to pull their $10.00 order from light pool.
However, the article wasn't clear enough for me to be certain.
Here's how you game them without being detected. Have your long term hedge fund/mutual fund department set up an connection to that market. Make sure that you only do long term investing on these venues (e.g., buy blue chip stocks that have low PE) so that they don't ban you.
Now you feed the information about displayed liquidity on high volatility stocks that you are interested in trading (e.g., small biotechs) from those exchanges. Especially this new one that isn't even a dark pool but has displayed liquidity.
Now next you use that information to front-run the mutual funds who are trying to execute their VWAP in other exchanges such as BATS. Most investors don't have access to dark pools and dark pools aren't obligated to conform to NBBO, so you could get cheaper shares elsewhere and sell when the VWAP is reported in the next hour/end of trading day by dark pools.
Other way you can play this is play the liquidity rebate game. So much rebate, $.14/100 shares. Just trade C all day long, high liquidity, low slippage. Offer and buy back at same price as long as C doesn't slip too much. Guess who's paying for the rebates, the mutual funds who's taking the liquidity. Wall Street, what a scam.
Now HFT algorithms, daytraders and market-makers all monitor this full-depth book for patterns whenever a large order is coming in. Suppose I'm a Fidelity Investment trying to acquire 5 million shares of MSFT for my $20 billion mutual fund, if I just put that huge order out there on the order book. Everyone else will jump in and buy MSFT; front-running me because stock market is like everything else, supply-and-demand; when there's a huge demand and you buy the supply ahead of time, you can charge more and make profit.
So to disguise my huge order, mutual funds prefer dark pools where the full-depth of books are not maintained. Instead, it's like shooting fish in the dark. A big mutual fund wants to buy 5 million shares of MSFT at 25, another big fund wants to sell 5 million shares at MSFT at 25. You submit your order to that market totally blind because there's no quote book and just have to see if your order gets filled. But the positive side is that no one could see where the market demand and supply is, so the mutual funds gets their orders filled and not front-runned. Dark pools also typically limit their participants to large institutional investors to limit the information because what HFT firms used to do is to "ping" the dark pools, send out random orders to buy 1 share of MSFT at a certain price to see if there's a "whale" order out there and then proceed to front-run. So lots of them got banned.
Now, the problem with dark pools is that there are sometimes not many people who trade on it because it's "dark," kind of like egg-and-chicken problem, no one could see the full-depth of book and so don't put their orders out there. That's what's called lack of "liquidity". Exchanges want to have people trade, because the higher volume, the more people will come and trade on that venue and the more money they make. So this "light pool" tries to fix this problem by having "displayed" quote book but the market participants are still limited and regulated still to make mutual fund participants comfortable about not being front-runned.
Well, as the saying goes, "it's never illegal unless you get caught"; so suppose if you are a huge fund or a bank with a HFT prop desk. You could register your mutual fund section with this "light pool" for its displayed liquidity. Now, you might even do some real trades on the exchange to make everything legit; but feed the market data section for your HFT prop desk which is registered with a totally different LLC designation, and have that fund execute orders on other exchanges based on information from this "light pool" (e.g., whale buy order on IBM on "light pool," front-run IBM on BATS).
Now for liquidity rebates on these exchanges, exchanges are now locked in a bitter battle to see who can attract the most volume and trades (because it's a snowball effect, more volume, more interested traders who come on to trade, more money). So they offer "liquidity rebates" for traders who put orders out there on the full quote-book because the more entries on a exchange's order book means that there are greater potential for greater volume on a exchange. A more concrete example is, let's say I'm a liquidity rebate trader on Citigroup (NYSE: C); C is last traded at $4.30, with national's best bid at 4.29 and national's best offer at 4.30 (NBBO). So I could submit out a sell order for C at $4.30 on the exchange's quotebook (I don't have to physically own C, I could short sell the stock). Now although NBB was at 4.29, someone might come alone and they are really impatient and submit a market order to buy C at $4.30 and they hit my order; now they are "taking" liquidity that I put out there on the market. So they are paying the exchange for the liquidity for doing so, typically, $0.02-$0.05/100 shares and to encourage more liquidity providers like me, the exchange passes some of that profit to me, $0.01-$0.03/100 shares.
Now a stock like C is heavily traded, there are literally thousands of orders on C's quotebook just between $4.30 and $4.29; suppose that a bad/good news come out on Citigroup, many of the traders who have their offers to buy or sell C at these cents increments might not all cancel their orders to adjust the valuation of C to the new news. This is called "low slippage," that in a high volatile event, you could trade out of your positions very easily with no "slippage" as trading in a small cap stock where in a volatile event, no one's willing to trade with you.
This is perfect for a liquidity rebate trader who literally goes around all day, offer to sell C at 4.30 and then buying back their short C shares at 4.30. They break even on the trade and get to collect $0.02-$0.05/100 shares liquidity rebates. So you do this over and over again on the market, generating higher volume on the exchanges and get to collect more rebates and everyone's happy at the expense of the liquidity takers. Now, with this exchange, the liquidity rebate is at $0.14/100 shares because they want to be attractive to the liquidity rebate crowd and generate lots of volume; and because the rebate is high, you could afford even higher slippage on C.
Take "e.g., buy blue chip stocks that have low PE) so that they don't ban you. .. Now you feed the information about displayed liquidity on high volatility stocks that you are interested in trading (e.g., small biotechs) from those exchanges.."
If you are trading blue chips at an illiquid (CSFB run) market center what information will that give you about "small biotechs"?
".. dark pools aren't obligated to conform to NBBO" : this is an incorrect understanding of how US stock markets work -- dark pools, with exceptions that don't count for too much volume, will have to print at the NBBO (Reg NMS, Rule 611, if you want chapter and verse http://taft.law.uc.edu/CCL/regNMS/rule611.html).
Also you don't need to trade at this venue to see what is being traded there: another provision of Reg NMS requires all ATSs to print within a short period of time into the tape, which is also by law disseminated nationwide as the last sale info in the name. By simply looking to see prints from this center, and which side of the midquote the prints are on (called the Lee-Ready test in the field, after a 1991 paper by those guys) will tell you if there are long duration buy orders or long duration sell orders active.
Turning to the merits of the proposal above, seems like a good idea, although the fact that csfb (or any big bank) is running it will cause most experienced traders to be skeptical of the exalted rationales being presented. If a broker that is handling flow runs a market center as well, without exception they wind up abusing the flow or the market center to make more $$. Have read a lot of good stuff here; keep up the good work, all.
Your second point of RegNMS requiring exchanges to print all their trades. Dark pool report their trades in their own idiosyncratic way. Suppose a large block order went down, I and my counterparty could agree on a VWAP price and just slowly complete our trade over the course of the day or even several days to obscure the order from the market. So reporting might be done, but it might be done as slow and as obscured as possible.
And talking about Credit Suisse, CSFB and other banks used to have a suite of VWAP algo's for their customers but they have curiously not updated the algo in the past 4 years. Why not? CSFB's internal trade desk probably is probably constantly developing better algo's but because they want to keep their own customers from having better market obfuscation and be able to front-run; they don't release these tools to their customers.
One of my big rules as a long-term investor is that I can't sweat the 1/8ths and 1/4ths (borrowed from Philip Fisher). The time I spend worrying about these high frequency traders getting a few extra cents out of me is time wasted finding great companies that are selling at a discount.
Sure... Its annoying, but if you really are a long-term investor a few tenths of a percent won't kill you.
I guess what I'm saying is that I don't care either way.
Just another way to shaft their own customers, isn't it?
no one is obliged to have an iphone or credit card too. that doesn't mean that a lot of people can't have them, and the fact that they got them voluntarily (or more likely trough carefully crafted advertisement) doesn't mean that companies providing them can now do as they wish with their customers!
believe they'll gain some advantage through doing it - so conversely you could argue that by offering this service Credit Suisse are doing their customers a favour
you're doing your customers a favor when you're actually giving them some advantage, not making them believe in you doing so.
Markets are based on trading. If there are no trades, there are no markets. If you want to buy 1000 shares of ABC company, and nobody has 1000 shares, guess what, the trade is not going to go through. This is what will happen on a restricted market.
Similarly, high-frequency trading means price corrections occur more quickly, meaning that when you buy or sell security foo, it is more likely at the correct price. Now you can argue that nobody really knows the correct price, but that is orthogonal. (Computers make mistakes, but so do people. There are some markets that are still not made on exchanges, and they are subject to the same whims that the equity markets are. Computers are buggy. People are irrational.)
My guess is that this market is for people with a lot of money that like to talk on the phone with bankers. They will get a "safe" investment (or so the dude on the phone says), and Credit Suisse will get a nice cut. Hint: whenever a bank invents a product, the main idea usually revolves around them getting a cut.
Replace "bank" with "just about anybody" and you have a winner.
Here's a better solution for us poor people: https://personal.vanguard.com/us/funds/snapshot?FundId=0085&...