Matthew,
Although I only have 3 trades, my slippage score not only stands out because it is in bright red but it also stands out because it shows an unwarranted -200.00%.
I use limit orders only. I always set them above the current ask to ensure that I won’t get a price that I am not willing to pay above the current ask. For example, yesterday the current ask for the stock I purchased was 36.94 and I entered a limit order to buy it at 37.00 or better. I got filled at 36.96. The nextday I see this bright red -200% slippage statistic.
Is the way I set my limit prices, to ensure that I get the price I want, going to seriously penalize me as I execute my trades this way all of the time or what?
As the explanatory text near the “Worst Case Slippage” statistic tries to caution, this is only one statistic among many. It is only a single data point and should not be obsessed about.
Regarding the stat: it has nothing to do with whether you use limit orders or not. It instead measures your average profit or loss versus the average size of the bid/ask spread in the instruments you trade.
Thus, a system that tries to scalp 0.03 cents profit or loss each trade, on instruments with a .05 cents bid/ask spread is asking for trouble. While you might be lucky to get that sort of limit fill, on occasion, most of your autotraders won’t, and so they won’t participate in the scalping profits, but rather will convert to market and get dinged by the size of the spread. The way to improve the stat is to increase the size of your profit versus the size of the spread. If you trade a stock with a five cent spread, you’ll probably want to aim for a trade profit higher than five cents per share.
The reason your particular number is so stark at this moment is that, first of all, you only have a tiny bit of data to look at (just a few trades); second, those few trades are unprofitable as a whole; and, finally, your P/L per trade is quite small versus the spread of the instruments you are trading.
It will normalize over time as there is more data to analyze.
Yes, yesterday’s trade was was not meant to be as it occurred.
It looked like a scalped attempt because I entered the trade about a minute before the market closed and then discovered that I was closed out of the trade 1 minute later at the market close because I selected the feature on my edit screen that tell it to close out all open trades at the end of the day.
I forgot that I had set this feature.
For the record, I am not a scalper of a few pennies as yesterday’s trade would imply.
PJ,
If you didn’t use limits would the system performance be that much different?
If so, then maybe the slippage stat is doing its job well, as it would imply your performance is only realised by getting that specific price and as MK points out your subscribers might not be so fortunate.
If not, then why use them? Tom Rehberger spent two decades designing trading systems for hedge funds and commented that stops and limits are two of the best ways to curve-fit a model to get results that will never be duplicated in real time.
Something to think about.
Broadsword,
There is one very important reason to use them and most people do not realize this but from my prolonged observations over the years of stock price behavior, I have come to the conclusion that market orders move price when placed, and limit order don’t. So, if you have a huge trade block enter a trade and you don’t want price to be immediately affected by the entry, you will be better off using a limit order. Even the smallest buy market order will make the price jump a few pennies or if its a sell market order the price will drop a few pennies. This rise or drop becomes even more magnified when the market order contains a large share count. But it is my observation that a limit order will not do this.
For this reason, I would discourage anyone from ever using a market order especially upon entry where you are trying to get the best price you can.
In a nutshell, we are talking about avoiding slippage here – Never use a market order for anything.
I would hope that my future subscribers follow my lead on this one because if they don’t then a large following of subscribers using market orders to follow my trades could (not necessarily would) adversely affect the intended consequences of my trade.
On the other hand, if I can get everyone to follow my lead exactly (which is why I would rather all my subscribers use autotrade so that I can be assured that no market orders are being applied at the time of my entry) then I am certain that everyone of them will be pleasantly pleased with the outcome of most of my trades – Just my opinion of course.
NEVER USE MARKET ORDERS!
Interesting observation. In my case I trade primarily futures, and I ALWAYS use market orders. For me it tends to correlate best to historical testing when I develop strategies (using limit orders has many pitfalls when testing new ideas, at least in Tradestation). And slippage is not that big of an issue on mini S&P (ES).
Thanks for sharing your experience and observations. It will hopefully get people to at least think harder about the way they do things, and maybe improve their trading.
I always like reading about how people do things differently than me - I think it makes be a better trader.
Kevin
I would like to add that use a market order is the equivalent of hanging a big red sign around your neck for all the market makers to see. You never want to give away your position. And the best way to fly under the market maker’s radar is to always use a limit order as these largely go unnoticed by the market maker who’s primary focus is the filling of market orders in an a way that can always adversely affect the initiator of the market order.
YOU WANT TO GET IN UNDER THE RADAR AND
YOU WANT TO GET OUT UNDER THE RADAR.
YOU CAN’T DO THIS WITH A MARKET ORDER.
The market maker can say “ah, look at all of these buy market order shares chasing the current price. Well since they all obviously expect the price to go up, why don’t I just temporarily bring it way down to shake them out, and afterwords, I will allow the price to rise as they correctly anticipated that it should but they won’t be in the trade any longer and I shall have taken all of those poor saps money because I saw them coming, ha ha ha”.
It’s certainly interesting to hear the other side of the argument. However I am of the opposite view. As a rule I always use try to use market orders especially for testing or building systems.
I think this ties in with discussions we’ve had before about how important (or not) people think entries are. For me I think way too much emphasis is put on them and it’s other factors that determine whether you make money or not, so if you’re going for large long term moves a few cents here or there on entry makes little difference to the overall profitability of a system but using a limit of course risks missing a trade altogether which can have a major effect, especially on a low win % trend following system with avg wins much larger than avg losses. Missing that one trade because you had a limit on the entry can be catastrophic if it ends up being a home run.
I didn’t see you mention though whether you’ve tested your system without using limits, it would be interesting to see if the results were vastly different. That would then tell you what it is your system is really making money from.
"I would like to add that use a market order is the equivalent of hanging a big red sign around your neck for all the market makers to see."
Again that’s interesting and although I can’t absolutely quantify it one way or the other I would have thought intuitively the opposite is true. Surely a limit order that remains in place until filled advertises your presence more than just completing the trade and walking away?
"Well since they all obviously expect the price to go up, why don’t I just temporarily bring it way down to shake them out, and afterwords, I will allow the price to rise as they correctly anticipated that it should but they won’t be in the trade any longer and I shall have taken all of those poor saps money because I saw them coming, ha ha ha"
Again, surely this would only be the case if having a specific favourable price is key to the success of their strategy, like if you were seeking to exploit price discrepancies (arbitrage) or seeking very quick profits with very tight stops, ie scalping. Anyone trading a longer-term strategy wouldn’t be stopped out by such a near term reaction. Maybe it’s true to say the shorter term the system the more important price limits might be to its success.
I real life, I have never experience not being able to get in because I used a limit order.
As I stated to Matthew above, I set my limit orders above the current bid/ask price to insure that I don’t get left out of the stock entry. Generally, this is all you need to do to make sure that you get in provided use are working with a stock that has exhibited adequate avg volume strength. Otherwise, just make sure that your share count upon entry is not exceeding the current bid/ask size as this is away that might prohibit you from getting in.
Finally, I think that the point that I was making in my post just above this one is the most important reason to use limit orders. This reason is more important to me than being concerned about a few pennies of slippage.
I should have emphasize this point rather than the slippage point as the most important reason why you should not use market orders.
Before electronic trading became available (pick your market: the start date is different for each, but the stock market has firmly been in the ‘available’ category for awhile now) limit orders had value.
Limit orders in a liquid market are a net loss in trading, even if scalping. If your market is liquid, a market order will make more money over time. Limit orders were created due to illiquid markets. That is their function. If you’re trading an illiquid market, I’d agree you would be foolish not to use one.
"Limit orders in a liquid market are a net loss in trading"
What does this mean? This makes no sense to me,
"If your market is liquid, a market order will make more money over time."
Where did you get that idea? Once you are in the trade, if the trend is up
for longs, your trade will make money just the same as a market order entry would.
As I stated before, regardless of the age we are in now, limit order still have great value in that they are largely ignored by the market maker. Market orders drive the daily price range of a stock and as that range is traversed during the trading day as the market maker fills the market orders, limit orders falling within that range get executed also. In other words, limit order executions are a consequence of the market order that was filled when it traversed the price vicinity of that particular limit order sitting in wait of execution.
For short term trading, being aware of the behavior of the daily price range is essential to its success. My system is short term trading system. So what I have had to say here primarily relates to 24-hour trading scenarios and it may not be relevant to the way that most of you trade. So, for most of you a market order entry is probably fine if you have no intention of exiting the trade the next day. But, for my system, flying under the market maker’s radar is a good practice to follow.
NEVER USE MARKET ORDERS!
this is a surprisingly uninformed thing to say. I have done a lot of entry testing. And there is a larger than expected number of entries, where the market takes off and never touches the limit. You are cutting out the best trades.
I would tell everyone to do the opposite of what this person says, unless trading illiquid markets. But then, I advise people to avoid illiquid markets. It is hard enough to execute trades profitably when also fighting trying to get decent exits or entries.
"his is a surprisingly uninformed thing to say. I have done a lot of entry testing. And there is a larger than expected number of entries, where the market takes off and never touches the limit."
No, it is not an uninformed statement.
I don’t know what kind of entry testing you are doing but you are missing the explanation of how I use limit orders to enter a trade. Obviously, the method of entry that you are testing and my method of entry are totally different.
I will simply state it again:
I set my buy limit just above the current ask price.
My limit orders are being entered in real time - not several hours or days before on a GTC order.
Do you understand now why your criticism is not valid?
Please read my entire explanations of my methods before you go off half-cocked. Ok? Your backtesting model obviously does not fit my system.
So please do not generalize, Ok?
If you’re not using a back-testable method, or trading illiquid markets my comments don’t apply.
My definition of liquid is: there is a bid and offer available 1 tick apart, with enough size to accommodate your trade size. For less liquid instruments, the same would be true if there is usually a size-appropriate bid/offer available at whatever the typical spread is for your instrument.
If your market is ‘liquid’ just search ‘limit orders’ on this forum for some great discussions over the years. Nothing has changed.
I have been playing around with various order types and the use or non-use of odd lots as well. I confirm that I have seen the same market movement that Bundle mentioned.
I have been using market orders on c2 b/c of the trade delays associated with c2. Using limit orders like Bundle mentions work consistently only if you are in direct contact with the market price (real-time feed). With c2 the order is 10sec or more old by the time your subs see it. To help with the delay, I set an instantaneous trade volume limit on entry to help ensure a consistent liquidity level, but limit order will just get converted to market order by c2 after 30sec anyway.
I actually trade what I call a "Smart Limit", which looks at the spread at 1sec intervals, adjusting the limit order every 10sec until it is filled or reaches some cancel criteria. Yes, hard to back test, but it does work very well.
Jim
DOH!!!
I will try this again. I quoted the 4 words you told everyone as what they should not do. I guess you missed my point, even though it was screamingly clear and accurate. I don’t give a whoopee about all your other explanations, because you are wrong.
I will make this much shorter so you might grasp it:
YOU: NEVER USE MARKET ORDERS!
ME: this is a surprisingly uninformed thing to say (rest left out so as to not confuse P. T. Barnum Strategies
If you need a further elaboration, I will try to say it in simpler words or with pictures.
DOH!!!
NO, YOU ARE WRONG!
If you are too closed minded to care what explanations I give in defense of my methods then don’t even bother giving your opinion at all because where my system is concerned, you don’t know what you are talking about.
You think you know it all don’t you? This is your problem.
When you hear something that you haven’t heard before and someone
comes along and tries to offer an alternative school of thought of the way
things really work you are too closed minded to listen.
I’ve been a student of stock market behavior since 1987.
How long have you been around?
Look, I’m done here.
I will let my system do my talking for me.
NO, YOU ARE WRONG!
DOH!!! CONVINCING PROOF!!! DOH!!!
If you are too closed minded to care what explanations I give in defense of my methods then don’t even bother giving your opinion at all because where my system is concerned, you don’t know what you are talking about.
You are seriously confused, as your system was never the issue. Your incorrect claim that everyone else should not be using Market Orders was the only issue.
You think you know it all don’t you? This is your problem.
You are confused again. That You know it all is the only problem. You asserted that everyone should not use Market Orders. Your bad. DOH!!!
When you hear something that you haven’t heard before and someone
comes along and tries to offer an alternative school of thought of the way
You were not correct, as a few other people tried to point out.
I’ve been a student of stock market behavior since 1987.
As a student, you didn’t learn much. But we have already established that with your anger toward market orders.
How long have you been around?
I made $5000 when the Oct 19, 1987 market dropped 23%.
Look, I’m done here.
You were done before you started
I will let my system do my talking for me.
That wasn’t the issue, I don’t give a whoop, and you will most likely be a distant memory in 6 months - you just don’t know it yet.
See you later Homer, DOH!!!
