Slippage is probably the enemy number one for short-term traders. What is slippage in live trading and how much slippage can you expect to have in live trading? What is slippage in real trading?
Slippage in trading is a hidden cost that is difficult to quantify. It’s the difference between a theoretical price and the price you get in live trading. In this article, we provide you with facts about slippage in live trading in a select choice of three different ETFs. Slippage is most likely lower than you realize.
First, let’s define what slippage in trading is:
What is slippage in trading?
Slippage is the difference between fictional results when backtesting strategies and the actual results in real life adjusting for commissions and transaction costs. It’s a “hidden” cost and is based on the transaction.
For example, a backtest might simulate an entry price of 38.09, but in live trading, you get a price of 38.1. This means you get a worse price in live trading (assuming you bought the asset). This difference might not sound much, but done many times, it can amount to significant amounts of money over a year.
Another example: If your backtest shows an entry on 100, in live trading this strategy might buy those shares at 100.02 – not 100. That means your strategy will be less profitable than when testing. This is slippage!
Slippage is not the same as commissions. Commissions are a cost we know. However, costs related to buying and selling are not always easy to measure. When you backtest a strategy, the entry and close are estimated on an executed price.
- Commission In Trading – What Is Realistic To Pay? (Definition, Example, Rebate, Slippage)
- Backtesting forex
Our experience tells us that the backtest results are always worse than real trading. But by how much?
It depends on the assets and strategies you are trading. SPY (S&P 500) has significantly less slippage than SIL (Silver Miners), for example, and breakout strategies are more costly to chase than mean reversion. It boils down to several factors. Thus, just using an arbitrary number like 0.1% or 0.05% doesn’t make much sense, in our opinion. Please read the link above to commissions in trading. We have established that we pay around 0.025% for a round trip in QQQ and SPY – including both commissions and slippage.
An example of slippage in trading – real trading
Let’s assume you want to buy shares in Apple. Apple might have a bid of 172.05 and an offer at 172.12. If you want to buy Apple you have to hit the offer at 172.12 or put in a lower bid. If you bid 172.07, for example, you risk not getting any shares unless someone hits your bid.
Slippage in trading is a hidden cost that a backtest can’t capture. When we backtest strategies, we always assume a negative slippage to allow for a margin of safety. We always expect live trading to have a negative slippage, but we also assume that the strategy gets worse over time.
Why is there slippage in trading?
There are mainly two factors determining the slippage in a trade:
Volume is the main determinant of slippage
The most important factor is probably volume. The more volume in an instrument, the tighter the spreads (the difference between the bid and the ask).
However, even high volume might not stop an asset from having high spread:
High volatility means more slippage
If an instrument or asset has high volatility, traders in the market offset this by bidding lower and offering higher. This means that the bid and offer diverge and thus you need to pay more if you want to buy and receive less if you are selling.
How can you measure slippage in trading?
There is only one method to measure slippage in trading: to compare executed prices in live trading to the historical data you perform backtests on.
This is a pretty boring task, but nevertheless very important for your strategy development.
The results in this article are based on daily bars: we go long or short with market orders at the close and we cover our positions n bars later at the close or open. We send our orders just a few seconds before the market closes or opens. We have described in a previous article how to enter and exit positions at the close. We use Amibroker and Tradestation to buy and sell at the close, and we go into code and details in our Amibroker course.
Thus, all the results are based on the price action in the last seconds before the close or in the first seconds of a trading day. This is official trading hours, not aftermarket or before the bell. Our broker is Interactive Brokers.
Slippage in XLP – live trading:
Trading XLP is not the most exciting thing in the world. Together with another ETF, XLU, both are probably the most boring trading vehicle there is. But in trading, boring is good. It might be more interesting day trading Tesla or other hot stocks, but we believe you are less likely to make any money compared to XLP trading.
XLP is a slow-moving ETF that consists of stocks like Procter&Gamble (PG), Coca-Cola (KO), Pepsi (PEP), Wal-Mart (WMT), Costco (COST), Mondelez (MDLZ), Philip Morris (PM), Altria (MO), Colgate (CL), and Estee-Lauder (EL). Because XLP has rarely any sudden and sharp moves, the spread between the bid and ask is always very low, normally 1 cent.
The XLP is a highly liquid ETF with more than 10 million shares traded daily. The high volume and the low volatility make the spread small at all times.
Live trading and slippage in XLP:
We looked at our executed prices in XLP and compared them to data in Norgate and free data from Yahoo!finance. We used our trading records for the last six months and we got the following results after about 50 executed trades:
Those 50 trades in XLP showed a positive slippage of about 1 cent. In other words, we get a better price than simulated in our backtests!
This was rather surprising, and not expected. A few outliers contribute to this, but overall, we got a better price in 40% of the trades. We expect this to gravitate toward zero or even slightly negative in the future.
Live trading and slippage in QQQ
Unfortunately, our results are not so positive for QQQ:
In QQQ, we have a negative slippage of on average 2 cents per trade, ie. 4 cents on a roundtrip. This is still a minuscule slippage of 0.01%, far away from ruining any strategy.
Live trading and slippage in SPY
Our third ETF for the day is SPY, the ETF that tracks the S&P 500, the oldest ETF around. Just like in QQQ we have a negative slippage:
In SPY, we have a negative slippage of 1.5 cents on average, 3 cents for a roundtrip. This is less than 0.01% per trade and has minimal impact on the results.
Ending remarks about slippage in live trading:
Picking the right assets to trade is important. Our trade records show that slippage is minuscule in highly liquid instruments like QQQ and SPY. For XLP, which is both liquid and slow, we had a positive slippage, although this was skewed by a few outliers.
All in all, slippage in live trading is pretty low if you stick to the most traded instruments and stick to your rules.
