mahkoh
1,065 posts
msg #129550
- Ignore mahkoh |
7/2/2016 8:47:19 AM
I believe the issue here is that VIXM has an average volume of less than 50000. The software considers the last traded price which does not necessarily reflect the bid or ask at the time the pair trade triggers a signal. VXZ does a good job at staying in sink with the bid/ask but you will not be able to get the requested price for VIXM.
|
Kevin_in_GA
4,599 posts
msg #129554
- Ignore Kevin_in_GA |
7/2/2016 4:26:45 PM
Agreed, but I check the bid-ask spread on VIXM several times a day and it is typically only $0.04, which pleasantly surprised me. Given that you are only trading 180-200 shares of VIXM per trade, the slippage is about $8 and the commissions another $2, so I usually attribute a loss of $10 per trade for VIXM and $19 for VXZ (since you are trading 800-900 shares at a bid-ask spread of 0.02).
|
mahkoh
1,065 posts
msg #129555
- Ignore mahkoh
modified |
7/2/2016 4:55:48 PM
Let's assume VIXM trades at 52.30 and VXZ at 10.80, resulting in a ratio of 4.84. Now VXZ moves to 10.60 and with an average volume of over 500000 it trades regularly. The bid/ask for VIXM has moved to 51.25/51.30 but as it has not traded since the move down the last price on Yahoo is still 52.30. The software reads a ratio of 4.93 and opens a trade.
Following the signal you can buy VXZ for 10.60 but who is going to pay you 52.30 for VIXM?
|
Kevin_in_GA
4,599 posts
msg #129574
- Ignore Kevin_in_GA |
7/3/2016 5:57:18 PM
Well, that is a challenge for any stock selected by any filter here on SF since they use the close for all calculations rather than the bid or ask. Without access to real-time data on the bid and ask, you can get stuck.
Looking at IB, there are issues shorting VIXM since they do not have a lot of shares available. However, you can use the same strategy on
SVXY:XIV (https://www.pairtradinglab.com/backtests/V3mHcQQHtGtd-Eeu) or
VIXY:VXX (https://www.pairtradinglab.com/backtests/V3mGxm040FIJCtLw)
with similar results. These four ETFs are highly liquid and likely not to be subject to large bid-ask spreads.
|
novacane32000
331 posts
msg #129576
- Ignore novacane32000 |
7/3/2016 7:13:01 PM
Any chance you offer this on Collective2 ?
|
StefanM
25 posts
msg #130020
- Ignore StefanM |
7/24/2016 8:27:03 AM
Hi,
I would be interested in trading these pairs, but I don`t know exactly how I could set up this technically. I am new to Stockfetcher and, although Kevin says that everything is determined in the formula (which is obvious) I can not read it completely. I understand that we wait for a z-score cross above/below 0, and I expect to see it here in Stockfetcher under the day.
But the Stockfetcher data are delayed, and so I looked at the Pair Trading Lab website. I entered the pair, but to be honest I do not understand what I see. Additionally, obviously Pair Trading Lab needs parameters for back testing and entering trades, and I am not sure how I should transfer the StockFetcher code to Pair Trading Lab.
I am not new to trading, but I am not familiar with back tesing and pairs trading. Nevertheless I would risk some money trading this strategy. I have read some of Kevin`s previous posts and I have no doubt in his understanding. I understand and agree that my question is not "fair" in the sense that I likely can not give back anything. But my question is if someone is in the mood to explain me the settings to trade this strategy through Pair Trading Lab. I mean, if someone would write me that it would take him one hour to set up this strategy and that I should pay for it a normal price then I would do this of course.
If you answer me please write slowly. I am handicapped not only through my non experience with back testing, but also because I write you from Germany :o).
I hope that I do not ask for something which is not usual in this forum. And now, may the force of the markets be with you!
|
ferndave
65 posts
msg #130096
- Ignore ferndave |
7/27/2016 5:56:49 PM
StefanM: Getting pair trades up and running isn't a few hours of how tos, plugging in the pairs Kevin posted, and you're printing money. Even with experience and extensive backtesting of "sure thing" pairs, you can still get burned trading live. I'd suggest spending more time at pairtradinglab.com. Read their Wiki. You don't have to understand everything, I still don't, but it is a very good background. Then start backtesting. Use the pairs Kevin posted or the ones that appear on the PTL homepage. Then play. The basic backtest settings aren't bad to start with. They often only need minor adjustments. What changes if you move the entry threshold from 2.0 to 1.5? Exponential to Simple? And so on. Better or worse win/loss or correlation or CAGR or commission fees?
Look at Kevin's VIX pairs. They are very closely correlated. Think about what other stocks/ETFs might do the same. Backtest them. Find pairs in the PTL database. Backtest them. I've run hundreds and hundreds and hundreds of backtests. You'll eventually learn what might have potential. I can backtest a pair and quickly have an idea if it is worth messing with the parameters or moving on. Note the pairs and settings that provide good results in a spreadsheet.
After all that, and you have 15-20-25 pairs that you think are solid, you can move to paper trading. Pretty easy through PTL's software. All the instructions are there. And then watch solid pairs in backtesting, be not so solid with real market data. Others will work fine. Then it's up to you if you want to go live and with what.
Let me know if something doesn't make sense in translation.
|
StefanM
25 posts
msg #130109
- Ignore StefanM |
7/28/2016 3:32:08 AM
fd,
thank you, that makes sense of course. My idea was that I have only a limited amount of time to look at this matter, and I thought about happily setting my trades and see what happens. In general I would expect that this pair might work for trading, but it is true that I do not really have an idea about the pros and cons of this strategy.
So I will look at this following your To-do list :-). May the force be with us potential future pairs trader! ;-)
|
ferndave
65 posts
msg #130122
- Ignore ferndave |
7/28/2016 10:51:16 AM
Stefan: I think this is a case where learning how to fish, or at least the different kinds of worms and fish, is a better strategy. I'm still learning or I wouldn't be using a paper account. Get your feet wet and ask questions. You'll get there.
|
Mactheriverrat
3,157 posts
msg #130149
- Ignore Mactheriverrat |
7/29/2016 2:03:44 AM
Nothing wrong in using paper account for as long as it takes.
|