Re: [JBookTrader] Stat for measuring bid/ask spread?

[Judson Wilson]

You would have to re-write a few sections of code, but it is not hard.
I have messed with this before, but I was just dabbling.

I would recommend adding it as a third (or fourth or fight0 parameter
to the backtest files, and then go from there.

Also of interest may be the 'book size', which is the total number of
bids and asks. I'm not sure how it pertains to the strategies you
mention, but I believe there is value in there, and this would be a
good time to add it as well if you are interested.

On Tue, Jul 31, 2012 at 11:42 PM, Lawrence Perepolkin
<lper...@gmail.com> wrote:
> Has anyone created a strategy that uses a statistic measuring the difference
> of bid and ask? I want to develop a mean reverting strategy that would use
> this. I believe when the bid ask spread expands the market is experiencing
> and increase in volatility. The strategy I am thinking of would use a option
> straddle or strangle to protect against large moves up and down and then
> implement at strategy in BookTrader to buy lower lows and sell higher highs
> in a tight range. In trend less time periods this works quite well. For
> example, after hours ES is mostly trend less, and one can make some money
> buying lower lows and selling higher highs. High volume trend less days are
> also quite profitable. On potentially high volatility time periods,
> earnings and Fed announcements, use option straddles/straggles no need for
> an ATS, on trend less days use and ATS for mean reversion trading about 100
> buy sell cycles a day. Has anyone considered this? Your thoughts?
>
> --
> You received this message because you are subscribed to the Google Groups
> "JBookTrader" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/jbooktrader/-/EOFMkhaOgX8J.
> To post to this group, send email to jbook...@googlegroups.com.
> To unsubscribe from this group, send email to
> jbooktrader...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/jbooktrader?hl=en.

[Judson Wilson]

That should read "(or fourth or fifth)"

[Eugene Kononov]

Has anyone created a strategy that uses a statistic measuring the difference of bid and ask? I want to develop a mean reverting strategy that would use this.  I believe when the bid ask spread expands the market is experiencing and increase in volatility.

Since you mention the ES, from what I've seen the bid/ask spread very rarely exceeds one tick during the regular trading hours. I have not analyzed the Fed days specifically, but what I'd expect is that while you may see the widening of the spread in the minutes surrounding the announcements, it would be very short-lived. 

[Lawrence Perepolkin]

A shorted-lived widening and an absence of liquidity is what I have noticed. I am not sure how accurate my programming would be in capturing this from book data. When I see this it can denote a minor change in a trading range. I am thinking that upon see this event, it may be a good time to enter the market and trade in a tight range. Following minor events of this sort seem safe, however the major ones, when volatility significantly increases is dangerous. I am not sure how one can avoid getting nuked by a flash crash.   

[Lawrence Perepolkin]

Judson, I was thinking of incorporating book size (bid/ask sizes)  as some kind of weight. I am digging into the code to see what I need to modify. I like booktrader's framework. Implementing all of its functionality in C++ would be a major undertaking.  

On Wednesday, 1 August 2012 09:57:22 UTC-7, Judson Wilson wrote:

That should read "(or fourth or fifth)"

On Wed, Aug 1, 2012 at 9:56 AM, Judson Wilson  wrote:
> You would have to re-write a few sections of code, but it is not hard.
> I have messed with this before, but I was just dabbling.
>
> I would recommend adding it as a third (or fourth or fight0 parameter
> to the backtest files, and then go from there.
>
> Also of interest may be the 'book size', which is the total number of
> bids and asks. I'm not sure how it pertains to the strategies you
> mention, but I believe there is value in there, and this would be a
> good time to add it as well if you are interested.
>
> On Tue, Jul 31, 2012 at 11:42 PM, Lawrence Perepolkin

>  wrote:
>> Has anyone created a strategy that uses a statistic measuring the difference
>> of bid and ask? I want to develop a mean reverting strategy that would use
>> this.  I believe when the bid ask spread expands the market is experiencing
>> and increase in volatility. The strategy I am thinking of would use a option
>> straddle or strangle to protect against large moves up and down and then
>> implement at strategy in BookTrader to buy lower lows and sell higher highs
>> in a tight range. In trend less time periods this works quite well.  For
>> example, after hours ES is mostly trend less, and one can make some money
>> buying lower lows and selling higher highs. High volume trend less days are
>> also quite profitable.  On potentially  high volatility  time periods,
>> earnings and Fed announcements, use option straddles/straggles no need for
>> an ATS, on trend less days use and ATS for mean reversion trading about 100
>> buy sell cycles a day.  Has anyone considered this? Your thoughts?
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "JBookTrader" group.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msg/jbooktrader/-/EOFMkhaOgX8J.
>> To post to this group, send email to jbook...@googlegroups.com.
>> To unsubscribe from this group, send email to

>> jbooktrader+unsubscribe@googlegroups.com.