Strategies to decrease slippage on trades

[almostthere]

I have been using Fidelity as brokerage. I have had an alarming amount of slippage on my buy orders both market orders and limit orders. The amount seems to hover around .015 (amount of slippage / price quoted at time of order on yahoo and Fidelity itself).

Now, I know the easy answer is 'use limit orders'. Frankly, I am getting similar slippage on those as well, so I am doing something else wrong.

I am currently trying to develop a strategy on spin offs. That said, I am buying about $500 of each spin off in order to gain experience in the area.

Things I could be doing wrong. Buying amounts of stock that are not multiples of 10. For example 2 shares of BIIB (about $260 each). I am guessing this takes me down the order queue to someone who is selling 102 shares, and I get their two.

I think another error may be placing orders at low volume periods, for example around lunch time.

Spins are volatile and that may be an issue, but most have very large daily volume when I am buying at least 1m to 3m shares.

Are HFT traders picking off my small orders for some reason that I can avoid?

If Fidelity just acting as the market maker on my orders and making a few pennies off me, and I need to switch brokers?

Feedback on what I may be doing wrong or what I could do better would be appreciated.

Any market studies that show the general times of day with higher and lower volume would also be appreciated.

[jacob]

You should define slippage in terms of ticks (generally 1 cent for stocks). What's the slippage relative to the spread?

Overall, you really shouldn't expect a budget operator like Fidelity to care overly much about getting great execution. Especially not if you're only trading $500 at a time. Keep in mind, you either have an actual real person on the other side executing your trades for you (and he is not going to waste many seconds on executing a $500 trade where he might earn a dime or less in commission) ... or it's going to get bunched together with a bunch of other trades and routed automagically.

It's quite conceivable that if you do a lot of order modifications, chasing the price with a limit order for example, you're just going to piss these guys off and they'll focus on other orders.

Also note that the US has about 40 different exchanges and darkpools for equity and that your order might be routed to any one of these. Reg NMS requires you to get the best national price at all times. This just means you wont execute unless it's actually the best price. However, if your broker mainly routes to the cheapest exchange (where they might get paid to execute), it may be that your order is just sitting then with all the actual trading actually going on at other exchanges or pools. Reg NMS also guarantees that your market order (once it shows up on the consolidated feed) WILL execute. There will be no slippage by construction (unless the opposite price you think you're seeing is not the actual opposite price---which seems quite likely if you're going by yahoo).

If you're trading commission free, expect execution to be rather shitty. You often get what you pay for.

Your yahoo feed is certainly not instantaneous (because that level of market information is expensive). If you want a 5 second delay, you can get a free one from BATS but that's just one exchange .. and it might not match the book on NASDAQ or NYSE or AMEX. The price can certainly move in 5 seconds.

HFT can pick off your orders. However, it's more likely that you'll see the price immediately move against you when you get filled. This is normal (even if was a human trader on the other side... humans pick off such orders as well). It happens because they know more than you and would therefore only go active if they have information contrary to your passive order. Ponder this until it sinks in if it isn't obvious already.

Round lots (100sh) used to be easier to trade but that was back in the floor trading days. It's no longer relevant. However, many people still trade mostly round lots so you're more likely to get filled completely if you stick to those. If you see 1 share being trades, it's likely a HFT looking for hidden orders (ice bergs, etc.).

PS: If you're attempting to construct some kind of trading strategy using tools available on the retail side, you should generally figure that you will always be the one paying the spread. Now, you can play the [spread] game at a professional level but this would generally require a co-located computer. IIRC, that'd run you about $1000-5000/month in hosting services.

[halfmoon]

HFT can pick off your orders. However, it's more likely that you'll see the price immediately move against you when you get filled. This is normal (even if was a human trader on the other side... humans pick off such orders as well). It happens because they know more than you and would therefore only go active if they have information contrary to your passive order. Ponder this until it sinks in if it isn't obvious already.

I did have to ponder this for a moment, then the light bulb came on. Should have been obvious, huh? Excellent reminder not to try shaving cents as an investment strategy, and also that trading is not always investing.

[cmonkey]

The scoundrels!

or it's going to get bunched together with a bunch of other trades and routed automagically.

I'm was surprised they would even pick off $500 orders, but then saw this.

[jacob]

@halfmoon - Trading is very different from investing which is again different from speculating which is also different from saving. It's possible but uncommon to do more than one at the same time. It's quite common for people to do one but believe they do another.

Trading is better compared to what the supermarket does when it gets a truck load of flour in the back ... and then parcels it out to customers at the front while scoring a profit for carrying the inventory, making the match, and keeping the doors open. Now imagine a bunch of intermediates (think of them as a bunch of independent contractors working in the supermarket) who are each willing to buy and sell depending on supply and demand at the backdoor and the frontdoor. Those matching the front door with the back door in a convenient fashion are the traders---and hence this is the stock market. Like with flour, you can get a better deal if you trade directly with the big guys, but you'd have to accept 50000 shares at a time. If you buy one pound at a time, expect to pay more per pound.

@cmonkey - The alternative is that the block order takes a lot longer to fill. TANSTAAFL ... people can have either free or almost free trades combined with 1 cent spreads and practically instant execution ... or they can go back to the old days where each trade cost $35 in commission, had a 25 cent spread, and took an hour to go through. In general, when you see people whining about HFT, it's old phone brokers who are going out of business, because most people prefer the new system over the old system. Of course, block traders run their own HFT in response. Here's the textbook stuff.. the real world is more sophisticated: https://www.amazon.com/Algorithmic-Trad ... 0956399207

[halfmoon]

@halfmoon - Trading is very different from investing which is again different from speculating which is also different from saving. It's possible but uncommon to do more than one at the same time. It's quite common for people to do one but believe they do another.

Now you're making me think harder.

My general M.O. is to purchase value-type stocks with reasonable dividends, 50% or lower dividend payout ratio and room for dividend growth. I make the trades by phone as a nod to DH's paranoia*. I try to buy rather large blocks at a time because of the higher phone commission, and I hold the stocks unless their P/E gets into an uncomfortable range and/or they do something crazy like suspending dividends. Am I trading, investing or speculating?

If this question is too elementary/vague, can you direct me to some sort of reading (below physicist level) that would enlighten me just enough to be dangerous?

*Extreme caution.

[almostthere]

First of all thank you all for the replies.

I can see that my first error was anchoring to a number that may not be as accurate indicator of reality, the yahoo last price number. The BATS top of the book and last ten trades is a more comprehensive view of reality from moment to moment.

The second error was not to be aware of the spread at any given moment and not to consider slippage as portion of the spread.

BTW Thanks you for the BATS idea. It is mesmerizing to watch the trade by trade data, and that is exactly what I did for the last issue I had purchased, BIVV, a recent spin from BIIB. I will be using it in the future to get a better feel for the market before placing any trade.

Yesterday I watched it for about ten minutes or so at about the same time as I bought it the day before. The spread was only about .04 or so. Most trades were in 100 lots with a 33 and 67 lots every once in awhile.

If the same were happening the day before (the volume was similar), the difference in my 11 share market order order was still .73 versus a spread that may have been .04. I am assuming bid ask at the time was something like 42.01/42.05 and the last price was something like 42.04 based on the data I saw yesterday. I then got immediately filled at 42.73. I realize I am making a big assumption that I could compare the day before to the next day and that ten minutes of watching day after was enough of a sample to accurately infer something about reality. If the same thing happens in the future though when I am considering the more comprehensive bats data, I'll need to do some serious rethinking about my account at Fidelity. If I am going to get bad fills at 7.95 per trade, I might well go back to Vanguard where I could bad fills for $2.

@halfmoon, it sounds like you may be relatively young in your investing career. I wouldn't worry too much about trading versus investing yet. The ideas at this point for you would seem to be diametrically opposed, and you would be likely to dismiss one or the other as untrue when they are both in fact relatively accurate models of reality. I hope you will not see me as speaking down to you. Even experienced market participants have a hard time holding both models in their heads and openly dismiss one or the other. I have only recently gone down the rabbit hole of studying trading, technical analysis, and trend following after 15 years of reading and interpreting markets from a fundamental and value perspective. It is cognitively very taxing to resolve the various models in my head. That said, it you want to know more start with Reminiscences of Stock Operator and Market Wizards. Also note, that when Jacob writes he is not only holding these two models in his head, he also uses another set of models that fall into a broad category of quantitative strategies. Everything he says is extremely perceptive, but you need more models to comprehend it.

PS Jacob - I finished one of the earnie chan books you recommended and have moved onto his second one. I haven't run his code yet, but plan too eventually. Do you think I could run his Matlab code in Octave? Otherwise, if I can improve my R quantstrat skills, I'll do it there. Many thanks for the reference.

[halfmoon]

@halfmoon, it sounds like you may be relatively young in your investing career. I wouldn't worry too much about trading versus investing yet. The ideas at this point for you would seem to be diametrically opposed, and you would be likely to dismiss one or the other as untrue when they are both in fact relatively accurate models of reality. I hope you will not see me as speaking down to you. Even experienced market participants have a hard time holding both models in their heads and openly dismiss one or the other.

@almostthere, I never mind learning something (as long as it doesn't make my brain hurt too much ). I'm an old fart in years, but most of my adult life was spent investing in aggressive-growth mutual funds. Only in recent years have I moved toward individual value/dividend stocks. It's true that my intent is to invest rather than trade, and I see the two simplistically as long-term relationships versus one-night stands. Time to go read the book you recommended. Thanks for the advice!

[almostthere]

@Halfmoon, brilliant then jump right in. There are many models out there and the more you have the more you can try on to different situations to see which fits.

[halfmoon]

@almostthere, I just picked up Market Wizards from the library. Thanks again for the recommendation!