I've been (automated) trading exclusively leveraged ETF over the last 4 years successfully. Obviously you need to make sure to pick a combination of ETFs that have high volume, high volatility (and ideally low corrolation). Personnaly, I trade UVXY, JNUG, RUSS and UGAZ and their inverse (instead of shorting).
Also to the comment about slippage, with high volume ETFs, the spread is usually quite thin. I've also run tests over a couple of months to compare market orders versus limit orders and when you consider the fees rebate and the fill price, the dynamic limit orders are better than market orders and as such is the only thing I now use.

Hi there, I saw your comment below, and wanted to ask you a quick question. I have been looking everywhere online and have not been able to find an answer... I would REALLY appreciate any advice you have:


I currently paper trade leveraged ETFs; like UVXY, SVXY, DUST, and NUGT. I have a decent patterned system that I have developed, that I hope will work in the real world.
I am concerned about slippage. Your post mentions it. I am also in Canada, and am using BMO's Market Pro software.
I am wondering about your fills. My system focuses on price patterns, capturing ~2-3 cents on a $20-25 ETF like UVXY, and ~5-15 cents on their more expensive (~$60) inverses; like SVXY and NUGT. I will plan to use a decent amount of buying power to capture these small movements. If my fills are very close to the posted asks and bids at the time I send my orders, everything should be fine. I am worried about excessive slippage with larger orders, and do not have the experience of having many orders filled to see for myself.
Both UVXY, SVXY and DUST and NUGT are very heavily traded ETF's, as you know. I was hoping that this would help liquidity and help prevent slippage.
What do you think?
Thanks in advance,

You will probably have to do some real tests to know for sure as there are quite a few variables involved:
- liquidity/market depth at the time of making the trade
- amount you trade
- platform used (I use Interactive Broker with NinjaTrader)

Historically I generally have only a few cents slippage on every trades (compared to paper trading) but I coded my own dynamic limit order to improve trade efficiency. Hope this helps, good luck.

Thanks for the quick reply. I guess I will have to test it out myself. It's nice to hear from someone with lots of experience with this, as I can't find any direct answers from either my brokerage, or online searches...

My system would be using ~2000-5000 shares of UVXY, and about 800-2000 shares of SVXY. As mentioned, it aims to capture ~1-3cents on a $20-25 ETF like UVXY, and ~5-15 cents on its more expensive inverse, SVXY. A slippage of a few cents on a UVXY order would probably effectively destroy my strategy, as 2 cents on the way in, and 2 cents on the way out would eat all my profitability...

Just so I understand you correctly, I'll offer a scenario; UVXY. Price is $20.29. Ask $20.30. Bid $20.28. Are you saying that if you submit say an order for 3,000 UVXY @ ask of $20.30, that you will get a fill at ~$20.31-33? And then if you sell @ bid of $20.28, you would get filled at ~$20.27-26??
How can people or algos scalp like this? slippage this bad would effectively destroy any scalping strategy other than market-making, would it not?

Thanks in advance,

(edited bc I messed up the quote numbers)

Profitable scalp trading with retail tools is indeed very difficult and based on what you describe, I doubt your algo will work live, at least based on my experience of developing/testing live similar kind of algos. BTW I was trading amounts in the same range than what you describe. Even to get this range of slippage (which was acceptable for my algos to be profitable), I had to be more "patient", not trying to get an immediate market fill but rather "dancing" with the market for few seconds/a minute.

My system is not driven by a machine algo, but by my own. I would be entering the trades as mkt orders, as my system is a pattern-driven one that cannot wait more than ~2 seconds for fill, and sending to mkt would offer me the fastest fill.
I'm trying to get a sense of what to expect.. I've spent all my time trying to get a handle on what the line is doing, after being told I could get timely execution by the broker. I had not started worrying about execution until recently, as I am about to go live.

Just to clarify: you think that I will not be able to get a fill of 2-5000 shares of UVXY at the shown ask or bid price, and you would expect slippage of a few cents on either side of a trade that size?
Thanks in advance,

Exact, unfortunately I don't think you will be able to get filled at the bid/ask with a market order, with that amount.
If you are trying to arbitrage the 3X vs the actual theoretical value (which I did try years ago), it worked on paper not live.

Damn, that's terrible... I was thinking I could get decent execution bc of how heavily traded those ETFs are. I also was told (or read somewhere) that the Market Makers are under obligation to sell or buy your shares for the posted ask/bid prices.
If I cannot get filed at those prices, my whole system will be out the window...

Any thoughts or experience with day trading a 2x or 3x's leveraged ETF. I have never tried this but its seems to me that even some of the 3x's ETF charts run a lot smoother than for example, CL or ES