Zerodha Part 2

Status
Not open for further replies.

Jai Mata Di

Well-Known Member
Supose, I have 20 lakh in my bank, but I kept only 5 Lakh in my trading a/c with Zerodha. I use that whole amount while trading, some for future and rest for option. And after 3 month my total capital in trading a/c is 25 Lakh, profit 20 Lakh. Then I have 400% profit but my bet size is 100% not 25%.

It was the actual case for that s/w developer and only a veteran & very high risk trader can bet 100% of his money. If market is range bound during those 3 month the result would have been different. ;)
 
Supose, I have 20 lakh in my bank, but I kept only 5 Lakh in my trading a/c with Zerodha. I use that whole amount while trading, some for future and rest for option. And after 3 month my total capital in trading a/c is 25 Lakh, profit 20 Lakh. Then I have 400% profit but my bet size is 100% not 25%.

It was the actual case for that s/w developer and only a veteran & very high risk trader can bet 100% of his money. If market is range bound during those 3 month the result would have been different. ;)
OK, so he made 100 % ONLY. Fine! Lets all try and beat that instead of cribbing!!!!
 
The limit of 200 lots or 10000 nifty per order is put by the exchange and unfortunately nothing that we as Zerodha can do about it.

That said, some of our high volume trading clients use our product which is called as " i3" , there around 11 order execution algos.

When I say order execution algos, it will give you the best way to execute an order(once you have decided what to buy/sell) with the least impact cost.

We have something called as a iceberg order, in the iceberg a big order is firstly hidden from marketwatch(similar to how in an iceberg, only a small bit of the iceberg you can see on top of the sea, and the 90% of iceberg is hidden in the water, ;) remember titanic) and also the order is broken into small parts and sent to execute. This not only gets over the limitation of the 200 lots, but more importantly will save you a lot of money in impact cost. Today if you put a market order of 10000 nifty, you could loose between 50 paise to 2 Rs in impact cost itself. This could mean 5000 to 20000 in impact cost saved. Brokerage actually is immaterial. We have a bunch of our high volume clients using this product.

The catch here though is that since these are algos, you will need a dealer terminal to execute and not possible on retail terminal. To get a dealer terminal, you need to become an Authorized person with us(cost around Rs 5000/exchange) , give us NISM certification and then also subscribe to i3( Rs 5000/month). But if you are a person who executes big orders, this would be peanuts compared to what you can save.

Will get Nithin to write a blog on this tomorrow..But here is the list of 11 algos available..

Aggressive execution strategies: Discretionary, Market/Limit if touched, Pegged to market, Sweep to fill

Reduce your risk: Bracket

Better average price: Mid point, active relative, passive relative, Scale

Automated slicing: Iceberg, time slicing, volume participation.

Which particular NISM exam needs to be passed and certification produced?

What are the requirements to become an "authorized" person, what is the 5000 (is it per month or per annum {its not a concern, just curiousity}?

Will I be able to open a PAMM style acct in this fashion to trade three different accts at the same time?

Is there anything else required besides 5000/month needed for i3 subscription?

Are their any minimum capital accts to get a dealer terminal. If so what are they?


On the data and software ...... for now i plan on using the amibroker plugin and gdfl data.

So let me get this straight,

a) when i place an order to buy nifty at 6010 on the chart, will the offer at that second on the ZT be 6010? i.e. are the two data sets (GDFL and ZT) in harmony?
(this should be moot, as it is a regulated exchange not frickin FX)

b) When the NSE disseminates tick by tick data. Are all the same price ticks bundled together or the same price ticks sent as one tick.
This could be the difference between a 70 tick bar being yay long or nay long of u feel me. In other words does GDFL or/and ZT use the TCP or UDP protocol? What does the NSE use? are they the same?

c) Sorry for this question, i i think i know the answer, but just to make sure, is Zerodha ever a counterparty to our order? (u mentioned DD).


d) IF we have a retail terminal, can we manually revise the SL (ie trailing)

e) Can we manually put in the TP or SL targets seperately, then is the order OCO when either executes or do we have to manually remove the other. In case of multiple entries, is it FIFO or LIFO.

f) I have patiently read 1300 pages of ur guys progress in the last two weeks, i was hazy on the above, also i started following u last August, and was surprised to see the breakeven on 10 lots drop to 1.37 points from around .74 pts a few mopnths earlier (in the past 6months or so), i cant see it but did u reduce something? Is 1.28 the cutoff? This is in addition to the spread. What is the avg spread for the nifty on ZT btw, for say the last month (if possible)?


Im trying to think of anything else (the ZT and GDFL data is an issue that is very important, i mean i execute a buy in the charts of the GDFL data but THEY WILL get executed in ZT (after confirming of course) in ZT data. same for SL TP.

Thanks for the help for the past couple of years. Im going to open a small acct in august and trade as of 2/9/13. After which i should hopefully scale up within a years time. Hence the dealer terminal question.

One more thing as an aside. I realize and u realize India is still pretty backwards trading wise and in other areas as well but that dosent mean cause some morons figured out how to curve fit data to past history, that backtesting or optimizing has aNY REAL WORTh. no computer program has ever been devised that can get u an edge in the retail market. As far as pointing out setups, maybe BUT elemEnts of discretion will creep in there. Basically u cant switch on ur comp and go to sleep, not in India, not in tokyo or London. Having said that, we still cannot supply constant Alternating current in this country (reasons are real and many), and as humans we are addicted to A.C. So Timepass or w/e can chill out, cause calling a place behind the times dosent mean u dont love it or the ppl tht reside thr. You can try and make it better. JaiHind.

To the dude who will eventually point out HFT programs, realize there is a acre or so complex in New jersey filled with servers, that compete to get the data first before all of us (the competition is so ****** up, that the companys who have servers away from the feeding point complained and the exchanges had to put in latency (minimal) on request to disseminate to the chosen ones fairly (horribly expensive too), So in effect these HFT's front run ur orders, which is all fine, cause we can still find our edge, IF we are smart. But yeah if u think u can put on an EA or algo and go bake a pizza, lol, think it one more time. Plus bactesting against historical data is a waste of time. Ill explain all this as i start to post in the non Zerodha sections.



The name is Vik. Sachin you and ur crew are good men. Reply when u have time, no rush. Im gonna make you famous.
 
We'd want the trader himself to explain this, Will get him to answer all the questions directly.

But one thing that I can tell you after seeing winners of 2 challenges( 1last year and 1 now), especially because I believe that there is always a method to the madness and love data analytic's, one common trait between winners which is missing among the people who have lost is bet sizing.

Most of the winners 70% of trades is just 1/5th of the maximum trade they will take, whereas the people who lose typically have just 1 trading size.

To give an example, if there is 1lk in the trading account 70% of the trades of winners is just 1lot. The remaining 30% they change it to 2 , 3 or more. These must be those trades where the change the size of the bet based on the conviction of the strategy.

The same thing if you ask most of the people who loose is that they have just 1 trading size for every trade. For eg if the account is 1lk they will trade with 5 lots every trade. When the account comes down to 50k, they trade with 3 lots every trade...

This was a common trait among the winners and hence sharing..

Anyone who varies his Position size probably does not have an idea of how an edge works in a market. When u have an 'edge' or can define an 'edge', u take all trades on ur signals regardless. The term " conviction of strat" or w/e is baloney.

By varying position size i do Not mean the actual lots, but more the actual percentage of acct risked.

Say u work on a 1/2 percent risk model, and scalp 10 points on a 10 pt stop a couple of times a day. Also u "expect" ur method to give u a 60% hit over a substantial sample, {{ for example 5 lakh acct can affor u 5 lots on the 1/2 percent model, 10 lots on 1% etx}}. Once u use fixed fractional then u STAY fixed fractional ............................. WHY????

Because u have no way of knowing which setup to cherry pick for ur risk o meter and apply different percentages too. (matter of fact a real shooter of a move will mostly never have an Ideal setup.... do u see why?)

If and only if u have a robust sustainable edge, then u TAKE THEM ALL, at the same level.

Seems like the challenge results are flawed. If i was a hedge fund recruiter/mgr/whatever, I would look at drawdown, average risk, average loss, average gain, max risk/loss/profit, Sharpe ratio, Max adverse excursion in trades tht won, hit rate etc. There are many ways to analyze traders performance, many metrics tht can be used, net profit is quite the worst. (yes just looking at P/L will drive u broke, any idiot can make a million dollars on one trade).


Her's the thing, when u buy nifty at a price, there as an equally smart dude on the other side trading on his own strat and timeframe who is willing to sell it. Even if a program (intitutional) buys at a level another program might sell there, mostly in the day price movement is 50 50 either way (small number of pts tht is), sometimes it is 60 -40 (when big institutions are in confluence), u strike then or hope to strike then. The funny thing is, depending on the timeframe u use, it might take u years sometimes to realize that you dont KNOW ANYTHING.

Trading --- the final frontier.

Im going to reccommend the first trading book i read

Nicky Taleb (Fooled by Randomnness)

Ur smart Sachin, friendly pointer, not being nitpicky (this and the algo post were the only posts i thought u were off in the past two years).
its very popular, most of u who have read it, could read it again, flipkart it, dont download it (respect and pay for knowledge)
 
That is scary...I thought around 25-30 % will be net profitable...but the %age looks a lot lower.

A good trader is also an asset to a broker...he makes money and he is a longer term source of revenue to the broker.

Final data is going to be very interesting ,that is for sure...

Smart_trade
I would put it at 5%

And after 5 years of consistent trading, 80% will get burnt out, and only 1% remain.

This does not include laymen, i mean so called pro traders or part time pro traders.

IMO./
 
sorry didn't realize your question earlier,

What you are seeing on trading platform or the data provided by data feed provider is typically which ticks every 1 second. It is not possible for you to catch every trade on the screen whatever the platform you are on.

So it is very easily possible that your trigger was hit and it was not shown on the data feed provider, nse or on the trading platform.
Ahhh this answers tht tick question. So its not true tick to tick, which is fine, but it is possible, just not here in India. Look into zen fire.
 

ryan4ucar

Well-Known Member
I would put it at 5%

And after 5 years of consistent trading, 80% will get burnt out, and only 1% remain.

This does not include laymen, i mean so called pro traders or part time pro traders.

IMO./
Many people on this forumn love to paint the negative picture always , to all negative people on traderji , stop crying and work hard on your trading skills only way to success in trading to to have small failures 1st then you improve on them , i dont know any successful trader who has not failed , if youll give up go to a corner and keep quiet dont infect other , sorry for the strong words but what i have said is true and painful to hear.
 
Status
Not open for further replies.