Wall Street’s preference for low-priced stocks

By Felix Salmon
June 13, 2012
Alex Tabarrok found an intriguing post by high-frequency trader Chris Stuccio.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Three weeks ago, Alex Tabarrok found an intriguing post by high-frequency trader Chris Stuccio. The idea is very elegant: if you want to stop high-frequency traders extracting rents from the market, there’s an easy way to do so — you just allow stocks to trade in increments of much less than a penny. Matt Levine puts it well: right now, he says, “because you can’t be outbid by another bidder within the same penny increment, you get free money by just getting there first”. If high-frequency traders could compete on price rather than just on speed, then a lot of the silly arms-race stuff would be replaced by better prices for investors.

It’s a serious proposal, so I was glad to see that Matthew Philips wrote it up for Businessweek. But after starting off well, Philips ends up joking about it, and refusing to adjudicate between Stucchio and his on-the-other-hand trader, Ben Van Vliet.

The fact is that on the face of things, Stuccio is undeniably correct. Here’s the chart, from Credit Suisse via Cardiff Garcia:

cs-24may12.jpg

The y-axis shows the bid-offer spread on any given stock, in basis points; the x-axis shows the price of the stock, in dollars. Clearly, there’s an artificial clustering around that curve. For a lot of stocks trading at less than $50 a share, the market would happily provide bid-offer spreads of less than a penny if it could; but it can’t. And when stocks get really cheap, the bid-offer spread becomes enormous. For instance, an eye-popping 3.766 billion shares of Citigroup were traded on December 17, 2009, when the stock fell 7.25% to $3.20. At that level, a one-penny bid-offer spread is equivalent to a whopping 31 basis points; if Apple traded at a 31bp spread, then its bid-offer spread would be almost $2.

Clearly, the traders were the big winners when Citi was trading at a very low dollar price — if you make the assumption that traders capture half the bid-offer spread on each trade, then the traders made almost $20 million trading Citigroup alone, in one day.

On the other hand, it seems that the market almost never trades stocks at a bid-offer spread much below 2bp. Which in turn means that for stocks over $50 per share, we’re pretty much already living in Stuccio’s ideal world, where the spread is determined by traders, rather than by an artificial rule barring increments of less than a penny.

Which brings me to my theory: that companies deliberately price their shares at less than $50, as a way of greasing their relationship with Wall Street a little bit. Back at the end of 2010, I was very confused by the fact that Facebook had done a 5-for-1 stock split, reducing its share price from about $75 to about $15. But in hindsight, maybe it was all part of its IPO preparations: you almost never see stocks go public at more than $50 per share.

Here’s a question for the data geeks out there: did nominal share prices decline after the stock market moved to penny pricing in 2000? If so, that would support my argument: that Wall Street manages to engineer stock prices so that it makes good money trading shares. And companies are generally happy to go along with Wall Street on this: most stocks trade at $50 per share or less.

It’s true that a lot of the rents from the sub-penny rule and low nominal share prices are captured not by Wall Street proper but rather by HFT shops. But all Wall Street banks have some kind of HFT operation of their own, and in general it’s probably fair to say that the lower the nominal share price, the more money that Wall Street makes. And conversely, the higher a company’s nominal share price, the less beholden it feels towards Wall Street.

To put it another way: the sub-penny rule is a way of allowing companies to price their stock so that Wall Street can make good money trading it. And we don’t need Stuccio’s rule, since it can effectively be implemented just by pricing your stock above $50 per share. If the companies don’t want to subsidize Wall Street, all they need to do is price their stock higher. And if companies do want to provide this hidden subsidy to Wall Street, maybe they should be allowed to do so.

Comments
24 comments so far

Felix,

Chris, Alex, Matt L and Matt P all fail to mention this, but there are already markets that do this: the European stock markets. All of them have a sliding-scale of price increments, so that the cheaper stocks trade in “sub-penny” increments. As might be expected, all the liquid stocks have spreads of 3-5 basis points, irrespective of the price. And, as Chris predicts, the penetration of HFT is lower there than in the US.

Posted by m_m | Report as abusive

Not in the USA pre/post decimalization , but James J. Angel’s Tick Size, Share Prices, and Stock Splits (
http://www.jstor.org/stable/10.2307/2329 494) documents exactly the relationship you suspect.

Posted by OneEyedMan | Report as abusive

Felix, I think you’re onto something.

Posted by Eericsonjr | Report as abusive

Whoa whoa felix, I think you are on the backwards train here:

“The idea is very elegant: if you want to stop high-frequency traders extracting rents from the market, there’s an easy way to do so — you just allow stocks to trade in increments of much less than a penny”

that’s just for LATENCY arb – but that’s just one part of HFT. To me, HFT is the high-speed automation of “normal” trading strategies: processing data (reading the tape, electronically, of course) and trying to deduce the future of stock prices from the price/volume action.

If you introduce more sub-penny price points, think about what that does: it gives the “smart” HFT algos, who are much better and faster at processing this information that we dumb humans are, MORE data points to aggregate: it makes it easier for the HFT Algo to figure out what you are trying to do. Sure – latency arb becomes less of a business, but Smart Algos dominate the world in this scenario. It’s the opposite of what the idiot retail monkey wants!

let me simplify:

if, currently, the inside market is 10.00 – 10.01 1000 x 1000, but with sub pennies, there were 100 shares offered at each of” 10.001, 10.002, 10.003, 10.004, etc… as you go to lift those offers, the HFT algos can respond, try to “figure out” what you are doing “hey, he just lifter 5 offers… I think he’s a buyer!!!” – and the cost to them to “obtain” this information is much lower due to the smaller spreads!

Many people complained that decimalization was the nexxus for the HFT revolution… subpennies would result in even more HFT dominance, in my opinion…

see?

Posted by KidDynamite | Report as abusive

sorry felix – one correction to my comment: subpenny pricing (narrower bid/ask spreads) would be great for retail monkeys trading tiny order quantities with no “information value” in them… just like decimals were a boon compared to teenies..

but subpenny pricing would make it much harder for institutions to execute orders of size without other traders (ie, smart HFT algos) figuring out what they are doing – that’s what I meant.

Posted by KidDynamite | Report as abusive

KD, point taken, which is kinda why it makes sense to just have share prices over $50, rather than subpenny pricing on share prices below $50.

But also, the algos used to enter large orders are very sophisticated these days, and are often designed to trick and even make money off precisely the algos you’re talking about. It’s a spy vs spy thing. Who would win is anybody’s guess, but in a way it doesn’t matter, since it’s a zero-sum game.

Posted by FelixSalmon | Report as abusive

This, taken further, is largely why Buffet hasn’t split BRKa, and why BRKb was issued at 3-digits (and never traded below that until the recent split — to a price well above $50).

Posted by dWj | Report as abusive

Er, Buffett, of course.

Posted by dWj | Report as abusive

** struggles to wrap room-temperature IQ around issue **

Umm, … if we just enacted a flat-tax of $0.005/share traded on all exchange trades, we’d kill HFT dead, make no difference at all to any genuine investor/trader, promote higher share prices (less price-reducing splits), and raise some revenue – which we kinda need, don’t we?

What’s that you say? – the Street won’t like losing the privilege of picking pockets for its daily bread? Guess that settles it.

** shakes head, mutters, walks away – in search of rope **

Posted by MrRFox | Report as abusive

There is, of course, the *much* simpler and *much* lower tech approach – quantize your timestamps, and randomly select from orders that occurs within a given quantum.
If we can have a penny as the smallest price-point quantum, there is no reason to not have a second as the smallest time quantum.
In short, all orders placed between 11:00:01 and 11:00:02 are of equal weight, with one chosen randomly. Same for 11:00:02 to 11:00:03, etc. etc.
This automagically puts a cap on the speed that you need (as long as you can get there within a given second, you’re good), and also eliminates a huge chunk of the wastage in HFT world (no, we don’t need Yet Another Fibre Linkup to reduce latency by 10 micro-seconds).

Posted by dieswaytoofast | Report as abusive

KD, as Felix points out, institutional traders are using algos too, and aren’t any more the helpless muppets of legend. Besides, even if their trades are more easily sniffed out, it is likely that the slippage they will face will still be less than the minimum of 1 penny they face now. As Chris opines, it’s a competitive market, and if HFTa and HFTb have both sniffed out the big trade, they can compete for the fill by offering a better price rather than being first in line.

Consider this: moving to decimal pricing from sixteenths would have had the same issues. Yet if you are willing to pay the same price with decimals as you would have paid with sixteenths (i.e, by wiping out 3-5 levels of the book), you have plenty of liquidity available to you: your trade will be done and dusted before the HFTs see the fills on the wire and can react.

Posted by m_m | Report as abusive

MrRFox: what is a “genuine trader”? Who in your world provides liquidity to the “genuine investor”?

Posted by m_m | Report as abusive

dieswaytoofast: simply imposing a time quantum will not do away (at least, not entirely) the advantage that higher speed can bring you. For example, if you allow dissemination of order-book state during that second, you will get close-to-the-wire sniping. If you don’t allow orderbook dissemination, the race will be to automate the collection and processing of every single source of news and information available anywhere, and which might have an impact on the price formation process. Once the technology exists, it is very hard, if not impossible, to go back to a world where the technology cannot bring you a benefit.

Posted by m_m | Report as abusive

** returns, rope in hand **
** slings rope over high limb of Buttonwood tree **
** fashions noose with rope-end **
** waits for passing big-time Wise Guy **

That the status quo forces are even acquiescent to the existence of an entire line of business whose sole ‘raison d’être’ is to do the kind of things KidD describes, ….

@M_M – a genuine trader is anyone who isn’t attempting to profit from front-running/pocketing the spread between existing bids and offers on the table and/or anyone who doesn’t have to worry about $0.0050/share being the difference between a profit and a loss. Could you step a little closer, please? – this rope won’t quite reach ….

Posted by MrRFox | Report as abusive

@M_M: I definitely don’t disagree with your liquidity points. In fact, I’ve used similar arguments in the past to explain why going to decimals didn’t “decrease” liquidity – it just spread it out more.

However, I would disagree on your point about HFT_a and HFT_b competing to offer liquidity once they “figured out” the order – on the contrary – they compete to TAKE liquidity – the liquidity that the institutional order is trying to take: that’s the kind of trading i’m talking about – you know, aka: trading: you buy what you think other people want to buy (this is what MrRFox is missing: that’s what TRADING is!)

Finally, while I just said that I agreed with you on liquidity when moving from teenies to decimals, I think there is definitely a point of diminishing returns with respect to liquidity, especially as liquidity can be removed at higher speeds as well…Smaller orders will continue to be happier, but larger orders will have to be smarter in concealing their information leakage.

Posted by KidDynamite | Report as abusive

@KidD – aha, not I know why my ears were burning.

“Trading” in my book involves deducing probable price action based on reported transaction prices for completed trades. Coring-the-apple by getting a look at unexecuted orders and “getting there first” by clipping part of the spread before someone else does is what “front-running” is in its essence. The former is gambling; the latter is even less savory – but it pays well, and that’s all that matters, right, Kid?

Not sure – this rope might be too thick for someone your ….

Posted by MrRFox | Report as abusive

I must be missing something. Whereas the bids and asks are in penny increments, the executed prices sure aren’t. The HFT’s are permitted to execute orders at increments not permitted to retail investors and others. This is where I see most HFT profits coming from, so why wouldn’t the answer simply be to level the playing field? Either other investors get to place orders in lesser increments than a penny OR no one gets to do so? My orders are constantly being shopped to the HFT’s at prices inside of a penny spread. How is that fair to the person whose limit order has been sitting there at the penny increment bid/ask who does not get executed because the HFT is coming in 1/100th of a cent inside the penny spread?

Posted by madridisburning | Report as abusive

KD: you understand all this well (I am a regular reader of your blog), but nevertheless it makes sense to put this down, if only because there is much misunderstanding about the nature and role of HFT.

I think honestly that the days of explicit “front running” algos as you describe them are long gone. At this point, the vast majority of HFTs are of two kinds:

- those which have (usually very short-term) “alpha”, driven by orderbook shape and flow, trade flow, inter-instrument and inter-market correlations, and newsflow. These are the equivalents of what you and MrRFox think of as “traders”; usually take liquidity; and so compete with your mutual fund trader.

- those which provide liquidity; they often have 0 or very small alpha; and often rely on the maker-taker fee model for the bulk of their profitability. These are the equivalents of the specialists and marketmakers of old.

There are some other exotics, which wade into dark pools or private broker-side liquidity pools (and thus give madridisburning his sub-penny trades), but it is my understanding that their number is small.

The slippage your mutual fund trader sees is usually because of a combination of 1 and 2: the aggressors see the flow of trades and, if their other inputs are in agreement, deduce that the market is going up and try to take some liquidity themselves; the liquidity providers come to the same conclusion and pull their orders. Buy-side execution algos have gotten fairly good at obfuscating their intentions, so it no longer makes much sense to deploy a strategy specifically targeting them; rather, the flow of trades is just one amongst a number of factors that the algos might consider.

With this sub-penny rule in place, the efficiency of this process is reduced: different algos may have different estimates for how much the price is likely to rise, but they are all forced to raise their offers by at least a penny. Similarly, if the price is indeed going up, the few passive executions to be had on the buy side will go to those at the front of the queue, so there is a race to get in as early as possible. Sub-penny pricing will mitigate much of this: algo developers will focus on improving the quality of their price estimates, so they can pull back or go in to the extent their estimates allow, rather than to the extent the rules force them to.

Posted by m_m | Report as abusive

fg

Posted by marketguru | Report as abusive

Felix, you clearly don’t understand how hft make money. They make money off of latency and frontrunning, not off of wide spreads.

The harsh reality of today’s market is that limit orders will NEVER get executed unless an hft machine is certain that the limit price is disadvantageous to the buyer. The hft can step in front by 0.0001 and take your liquidity otherwise. It doesn’t matter what your limit price is, as a slow trader you get screwed. If you set an advantageous limit, no fill, and a disadvantageous limit, instant execution at the worst possible price.

You also chopped off the second graphic from the Credit Suisse report: the one show the average size of the NBBO drops to 100 shares for higher priced stocks. It’s a fallacy (pushed really hard by the hft lobby) to identify tight spreads with liquidity. Repeat three times: tight spreads are not the same as liquidity, tight spreads are not the same as liquidity, tight spreads are not the same as liquidity

Posted by marketguru | Report as abusive

Over $100 it’s harder for a smaller investor to calibrate the size of their position, ie at $150 you can have a $900 position or $1050 but not $1000.

Options are in lots of 100 shares, so 1 contract is for $15000 of stock.

There’s a stigma in a low-price stock, some old-school funds and institutions have docs that forbid them, to the point where other people started ‘low-price stock’ funds because they thought there was a performance anomaly.

It’s an interesting natural experiment in the impact of tick size on bid-ask, but the sweet spot is around $50 because it’s always been there and market culture has grown up around that, so it’s best for liquidity and to have room to rise or fall before a split becomes desirable, somewhere above $100 or below $10.

It would be interesting to look at whether going from 1/8s to decimal pricing led to more low-priced stocks, which one would expect if market-makers were influencing pricing conventions to widen spreads. If anything there seem to be more high-priced stocks like BRK and GOOG, where the issuer is saying they don’t care about liquidity, they want to discourage short-term holders.

Posted by streeteye | Report as abusive

Here’s an even better idea….tax all trades held for less than a single trading day at 95%….that would bring an end to high-frequency trading and restore some much needed sanity to the markets.

High-frerquency trading is nothing more than gambling, and contributes nothing to the stability or the markets or to society in the real world.

Posted by mfw13 | Report as abusive

@marketguru..I agree with you completely. I don’t think the right issues are being discussed here. If I could step inside a ton of asks at the penny increment by beating them by .0001, especially when there are few bids at the penny increment below, AND have the order shopped to me, to boot, I too could make plenty of money. But, no one has offered me the opportunity because I am low in the food chain.

Posted by madridisburning | Report as abusive

@madridisburning,

You are so right. No one in the media, not Felix, no one, understands this. It’s criminal how bad it is. Only when you trade stocks and pay $1000s per year in “tolls” to the hft, then you understand.

There is simply no escaping. You can’t use limit orders any more than you can use market orders, despite the bull$— the hft promoters peddle. The best you can do is use a limit that reaches across a one penny spread and hope you get filled for a $0.0049 fee per share to the hft on every trade. (half of the spread, minus $0.0001)

Posted by marketguru | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/