Trading with Market Statistics VI. Scaling In and Risk Tolerance

[dbntina]

Good point Nick...yes that volume is included in the price...if it doesn't keep track of all of the volume at every price it wouldn't be accurate...good point.

We don't check to see if its the PVP until after we have stored the current volume at this tick at the appropriate array.

It's as accurate as you can get...no approximations here....it is adding up tick volume at every price...so as long as the tick volume is accurate the PVP is accurate.

My code was checking every single volume at every price in the array...after every tick...no need to do that...you simply check the current tick volume at the current price against the current pvp, if it is greater its the new pvp, if not then keep the old pvp...simple but effective.

Hope this helps you understand what the code is doing.

Good job Blowfish,

dbntina

[nickm001]

understand now... very nice improvement.
Here is another suggestion, if you have not implemented yet... store on every tick... check for new PVP only if the price changed... this is 60% savings and no loss of accuracy. (my initial concern converted into improvement )

[dbntina]

Nick,

We will see how it works next week.

Also, tried to run the VWAP and SD bands on a tick by tick basis and it won't run...way too memory intensive. There is no way around that I am pretty sure. To get tick by tick accuracy you have to go through every tick every time there is a new tick to check the difference between the new vwap and the price at every tick and then perform all of the calcs...just not going to work. Code is easy but my computer with 4 gigs of ram just won't run it...

later,

dbntina

[BlowFish]

	Quote:	nickm001
	understand now... very nice improvement. Here is another suggestion, if you have not implemented yet... store on every tick... check for new PVP only if the price changed... this is 60% savings and no loss of accuracy. (my initial concern converted into improvement )

Hi nick, this would not be needed. There is a single comparison on each loop...is the volume at this level greater than pvp.

Store this ticks volume
Is the stored volume > than the current PvP
No Finish

I think I am safe in saying the code is optimal. You'll see when Dbntina publishes.

Cheers,
Blowish (another Nick)

[NAVEEVIa]

	Quote:	jperl
	If you have followed the details for scaling in, you should be able to answer the following questions. Take a shot at them and post your answers here 1)Assume your account is 50K and that you have a risk tolerance on any trade sequence of 2% of your account including commissions. A trade sequence is a group of related trades involving scale-ins and scale-outs and/or reversals (which we haven't discussed yet). Let's say you trade the Emini Russell 2000 index futures where 1 tick is worth $10 (0.1 points) and your roundtrip commission/contract is $5.00. You enter a 1 contract trade long at the 2nd SD, but the market moves against you back down to the 1st SD where you scale-in another contract long. The market is relentless and it continues to move down to the VWAP where again you scale-in an additional 2 contracts long. Again the market continues it relentless move down. You finally hit your risk tolerance at the 1st SD below the VWAP, so you exit your entire trade and are flat. Question: What is the value of the SD in ticks? 2)When you scale-in at the VWAP in the above scenario, the market finally rotates and starts moving back up. You exit the entire trade at break even +1 tick. Question: Where is your break even point including commission measured in ticks above the VWAP? How much money did you make on the trade? If you were able to answere these questions without too much difficulty, then you are ready to become a full time trader with all the rights and privileges granted thereto. You are also ready for part VII

I wil try on this , plz correct me if i am wrong
Ans 1

Risk tolerance =2%*50000=1000
let SD be x ticks
1st entry long at 2nd SD , market moves down to 1st SD
so loss = 1contract * x=xticks
2nd entry scale in 1 contract at 1Sd, market moves down to VWAP
so loss =2contract* x =2x ticks, total = xticks above + 2x =3x ticks
3rd entry scale in 2 contracts at VWAP, market moves down to 1sd below VWAP
so loss=4contract*xticks=4x ticks , total=3x tciks above + 4x=7x ticks
at this pint risk tolerance hits so trader sells his 4 contracts.
total commisions=4contracts * 5=20$
risk tolerance =7x ticks loss + 20$
1000=7x*10(bcoz i tick is 10$) +20
x=14, so SD is 14ticks

Ans 2
Let breakeven be y ticks above VWAP
at vWAP previous loss is 3x ticks
at y ticks above VWAP his profit is 4y ticks
for breakeven
4y*10=3x*10+4*5
substituting x=14
40y=420+20=440

40y=420
so y =11 ticks
so breakeven is at 11 ticks above VWAP

[mmaker]

Scaling in .... this is exactly what i did at the roulette table in the casino while on my honeymoon.

Playing red and black or odd and even i doubled my bet when i lost.

I started with a $1 chip every sequence. If i lost i doubled. If i won i started the next bet with $1.

Boring most of the time except when i got up to $512 bets just to get my original dollar back.

I played 4 nights this way. I played for about 3 hours and a bit each night. Each night i left with $350 more than what i came with - even after replenishing my wife's purse.

They changed the guy rolling the ball quite a bit and they watched me real good but i stuck to my system and did not deviate from the rules i set for myself.

When i got back home from my honeymoon i looked up the notes for the system i thought i was playing and realized this was not the system my dad had shown me.

That was 20 years ago and i have only visited a casino twice since then. One was a charity casino where they were trying to take your money at odds that were worse than vegas and once with a friend because he wanted to play slots and i had never done that - so i put $20 bucks in one machine. Then i stood and watched him win $300.

In retrospect doubling your money when you lose in a casino is a high risk game. One long run will wipe you out. So you need a good grubstake if you take 20+ losses in a row.

I do not recommend it - too hard on the heart lol.

[mmaker]

For the cme (e.g. emini ES)

One can lease a seat if you are doing 25 round turns a day and save a considerable amount.

25 round turns per day = $24.5 at $0.98 per round trip (incl. clear fee $0.78 and globex fee $0.20)
lease cost per day = $58.4 (not sure about this number i think it is cheaper but site is down right now) ($58.4 times 252 days = $14,716)
one time application fee $1750
this option as is works out to $14,716 for lease plus $6174 in comm plus $1750 application fee
=$22,640

compared to retail cost at ib $4.8 times 25 round trips a day times 252 days
= $30,240

If you do more round trips the difference is even bigger.

If you don't do 25 per day or 1500 per quarter you pay retail.

You must incorporate to lease a seat or trade under a corporation.

http://www.cme.com/clearing/clr/fees/rule106r.html

You don't hear eTrade flogging this kind of commercial on TV.

[mmaker]

Jerry, MAE of 6.43 today on multiple contracts!!!!!

You take my breath away....

But i guess if you got 'em use 'em.

Actually there are some big traders around who risk $50k per trade. I think they do it on more contracts but tighter stops probably.

[jperl]

	Quote:	mmaker
	Jerry, MAE of 6.43 today on multiple contracts!!!!! You take my breath away.... But i guess if you got 'em use 'em. Actually there are some big traders around who risk $50k per trade. I think they do it on more contracts but tighter stops probably.

It really comes down to your risk tolerance. Once you start trading using risk tolerance rather than stops, you don't have a problem scaling-in multiple contracts. On days like yesterday, it was quite profitable.

[MidKnight]

	Quote:	BlowFish
	This chart beautifully illustrates the behaviour that I was talking about (and trying to ensure I understood) a few posts back. This mornings DAX again. From the session open it just dropped with no base to build any sort of volume for the PvP. Volume increased on the fall and so PvP led the Vwap the whole way down.

Hi blowfish,

Just an idea - but the DAX futures are open 1 hour before the cash. That is why the volume radically increases - it is getting closer to the open which then opens with a flurry of volume.

With kind regards,
MK