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)

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

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.

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

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.

“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 ….

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.

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 ….

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.