TopstepTrader and Patak Trading Partners- Any and all questions answered here

Status
Not open for further replies.
Quote from Maverick74:

Here, let's do this. Let's create a risk profile for every basket of 10 combine traders. Let's say they pay an avg of $200 each. Let's say one out of 10 actually gets lucky and goes live and loses 4k. And then goes live again and loses 4k and quits. So for the 10 combines they make 10 X $200 or $2000. Let's say 4 of those guys get a refund and take their money and run. So that $2000 is now only $1200. They lose 8k on that one bad trader who got lucky and had to eat his losses and they have 1200 from the combines to offset that for a net loss of 6800 for every 10 traders. We can change these numbers if you like.

Let's say only 1 in 20 goes live. And let's say instead of 40% getting a refund, let's say 20% do. So that's 16 X $200 or 3200. And let's say they are only eating 4k in losses. Even then they are losing $800 net for every 20 traders.
More...

Your first para assumes 10% first time combiners succeed. Thats not going to happen. Numbers in 2nd paragraph are a bit more realistic. I will work with numbers in 2nd para as a base case. In actual operation, I think 97%-99% combines will fail to move to live. But lets be conservative from the business owner point of view and lets agree with your numbers that 5% actually are able to pass combines in the first time.

You are off in the assumption that you are charging them 200$ but losing traders are costing you 4k. No, as per the current model, losing traders have 500$ daily limit and they get sent back to combines at 2x that limit = 1000$. So, that changes your calculation by a factor of 4. Please note that we are discussing everything in the context of a 50k account currently - since that costs you 200$.

So, out of 20, 1 pass. Assume 20% refund. So, 200*16 = 3200. Loss = 1k. Net profit = 2200$. If you can get 200 first time combiners every month, your net profit pm = 2200*10 = 22k. And this is only combine income. Off the guys who trade live, you also comms. Yes you have overheads etc.

Things become interesting, if you have a 400 first time new combiners every month. PL = 44k per month.

Above example is just analyzing the current model. Say, you do your combines for a longer period, and so the traders you get have a much higher chance of being successful. If you can produce 5-10 out of every 100 live traders to be good, and you get a good profit split.

Mav, I have done all these calcs before in the thread. You ought to really read everything "again" with an open mind. I know apart from this Patak episode, you are otherwise an intelligent guy. I can bet that you can design a good competitive product in the market.
 
Quote from Maverick74:

Here, let's do this. Let's create a risk profile for every basket of 10 combine traders. Let's say they pay an avg of $200 each. Let's say one out of 10 actually gets lucky and goes live and loses 4k. And then goes live again and loses 4k and quits. So for the 10 combines they make 10 X $200 or $2000. Let's say 4 of those guys get a refund and take their money and run. So that $2000 is now only $1200. They lose 8k on that one bad trader who got lucky and had to eat his losses and they have 1200 from the combines to offset that for a net loss of 6800 for every 10 traders. We can change these numbers if you like.

Let's say only 1 in 20 goes live. And let's say instead of 40% getting a refund, let's say 20% do. So that's 16 X $200 or 3200. And let's say they are only eating 4k in losses. Even then they are losing $800 net for every 20 traders.

See the only variable that makes your argument is if that guy who goes live actually makes money instead of loses money. Because that guy represents a Asymmetric payoff in that his earnings are unlimited. There is a positive slope there that can generate potentially huge returns. But you and I both agree that is probably NOT happening. So the numbers I gave are fair. I honestly don't see how they are much above breakeven.

And this ignores all the overhead cost as well. I'm writing that off as free for the time being to make this easier.
More...

See the deal killer here is the eating of the losses. That's what throws all these numbers off. If you get long this trade, you have short put exposure on the losses. You can't really quantify this very accurately. I mean you could have a really ugly streak where tons of guys get lucky and go live and you could eat 100k to 200k in losses as an outlier and only have 10k in combine revenue to offset it. Yes, I agree that is a 5 sigma event, but if one wants to seriously look at this trade, he has to accept that risk and I'm not willing to accept that uncertainty. One hedge I can think of is not allowing any refunds which would at least allow me to model that variable with absolute certainty.
 
Quote from traderchi128:

I have trained guys at my last firm while I traded....a long time ago. Is actually a trip helping new guys learn the biz. You have to have patience though because a majority of new traders, especially these days, are stubborn and don't understand what it takes to make it in the biz.

I'll still help new guys who ask me for info/insights on trading....but running a company. Nah. Not my thing. Too busy trading.
More...

traderchi- agreed! It is a great experience working with traders especially those that are truly open to learning..

Very cool you did this and gave back to the industry. Many think you aren't any good if you are coaching traders or teaching. I was taught similar to how you taught guys. Having someone to talk to that knows the business and points you in the right direction each day is so very valuable but as you know it takes TIME to learn.. and you never stop learning even after you get there. ha

mp
 
Quote from gmst:

Your first para assumes 10% first time combiners succeed. Thats not going to happen. Numbers in 2nd paragraph are a bit more realistic. I will work with numbers in 2nd para as a base case. In actual operation, I think 97%-99% combines will fail to move to live. But lets be conservative from the business owner point of view and lets agree with your numbers that 5% actually are able to pass combines in the first time.

You are off in the assumption that you are charging them 200$ but losing traders are costing you 4k. No, as per the current model, losing traders have 500$ daily limit and they get sent back to combines at 2x that limit = 1000$. So, that changes your calculation by a factor of 4. Please note that we are discussing everything in the context of a 50k account currently - since that costs you 200$.

So, out of 20, 1 pass. Assume 20% refund. So, 200*16 = 3200. Loss = 1k. Net profit = 2200$. If you can get 200 first time combiners every month, your net profit pm = 2200*10 = 22k. And this is only combine income. Off the guys who trade live, you also comms. Yes you have overheads etc.

Things become interesting, if you have a 400 first time new combiners every month. PL = 44k per month.

Above example is just analyzing the current model. Say, you do your combines for a longer period, and so the traders you get have a much higher chance of being successful. If you can produce 5-10 out of every 100 live traders to be good, and you get a good profit split.

Mav, I have done all these calcs before in the thread. You ought to really read everything "again" with an open mind. I know apart from this Patak episode, you are otherwise an intelligent guy. I can bet that you can design a good competitive product in the market.
More...

No, you are off as well. That one trader who went live and lost 1k went back to the combine and went live again and say he lost 1k again. Very reasonable no? So even with your numbers, we are only producing a measly $200 in profit (2200-2000) for every 20 traders not counting the over head. Again, this is not adding up. It's losses that ruin the attractiveness of this for me. And to make matters worse they are splitting that $200 monthly profit among about 10 people!!!! Wow $20 per person. I rather open a Subway Franchise. LOL.

Correction: That's 1200 in net profit, not 200. (3200-2000). Sorry.

Of course he could go back a third time and lose another 1k bringing up the losses to 3k total with 3200 in revenue to offset that.
 
in case it was lost..

Michael... I tell you what I'll do, I'll take you up on the free offer but it has to be for the 2k risk limit per day combine but I'll only take it when Ninjatrader is ready. You can post my dollar results too but not my trades. 10 day is fine.

I do agree with GMST that its setup as a scheme to generate combine fees.

Mav or anyone else who is saying you can't set this up to generate money.. sure you can.

Just look at a distribution graph to understand this...

Losers
Big Losers
Make a mistake/hit loss limit
Small Losers
Winners that break a rule/make mistake
=> $$$ PROFIT FROM ALL ABOVE
Winners
Big winners
Solid Winners
=> Break even on these "trades"
Huge Winners (3 std dev)
=> Small payout


You can adjust the difficulty, add silly rules (like they have), or what you have. Now at end of month, you just tally up your profits and fund enough traders from your combine profits. Too many pass? No problem. You've enough silly rules to find something wrong with most of them.. sit down.. have a pow wow talk about how they need to do better but show promise.. free combine. Of course then there are the upsales... room.. squak.. membership fees. etc

Why make a lot of rules for trader getting his refund? If combine fees aren't the revenue generation method then why even charge it? Mav, that's why I made a point to show that you don't get refunded even if profitable -- they added more strings so they can keep your money. It is a matter of principle and shows what the true nature of operation is.

Why not make it so that the trader has to trade a least 80% of days out of 3 months time? Make it free. How many will trade 80% of days for 3 months? You don't need to charge a combine fee to find serious traders.

I think its a great business for the combine owners.. Even in the "legit" form its a good business. and then if you are really sophisticated you can run data mining operatoins on the traders silently in the background... gather unique quant data for automating any profitable traders you find... No need to tell about the owners.. if one happens to be phd quant.
 
Oh god this is hilarious. Michael is willing to give it away for FREE and Lucias still doesn't like it. I think the future of this country is f*cked. Better start learning mandarin you guys.
 
Quote from Maverick74:

No, you are off as well. That one trader who went live and lost 1k went back to the combine and went live again and say he lost 1k again. Very reasonable no? So even with your numbers, we are only producing a measly $200 in profit (2200-2000) for every 20 traders not counting the over head. Again, this is not adding up. It's losses that ruin the attractiveness of this for me. And to make matters worse they are splitting that $200 monthly profit among about 10 people!!!! Wow $20 per person. I rather open a Subway Franchise. LOL.
More...

No, No, No! Incredibly few will make it combine-live-combine-live-combine-live. To be precise 2 out of 10000, assuming a 1%, 10%, 20% success rate for first time and 2nd timer live-combine cycle and 3rd time live-combine combiners. You can make your assumptions here and set up an excel sheet to change the numbers so that you can model different scenarios. I did a detailed post on this.

The correct way to model this problem is to separate it into 2 parts:
1) First time combiners risk:return and PL modeling
2) Combiners who were earlier Live and then went back to combines. Subsequent time combiners risk:return and PL modeling

As I showed in one of my earlier posts, very very very few people will actually fall in part 2)
 
Hey Lucias, what if Michael pays you to do the combine. And no matter how much money you lose, he backs you with a million dollars and he cuts you off once you lose 500k. And then as a parting gift, he pays you again for taking the time to try. Because you know, you did try after all, that's worth something. LOL.
 
Quote from gmst:

No, No, No! Incredibly few will make it combine-live-combine-live-combine-live. To be precise 2 out of 10000, assuming a 1%, 10%, 20% success rate for first time and 2nd timer live-combine cycle and 3rd time live-combine combiners. You can make your assumptions here and set up an excel sheet to change the numbers so that you can model different scenarios. I did a detailed post on this.

The correct way to model this problem is to separate it into 2 parts:
1) First time combiners risk:return and PL modeling
2) Combiners who were earlier Live and then went back to combines. Subsequent time combiners risk:return and PL modeling

As I showed in one of my earlier posts, very very very few people will actually fall in part 2)
More...

I disagree, I get e-mails everyday in my inbox when a guy goes live. They are backing one to two new guys a week. Your "assumption" is false. And I agree, most of those guys are probably just lucky and not good and that is the real problem here. Those lucky guys generate unlucky losses for the owner. You are making way too many assumptions. As a trader, you need to err on the side of being wrong, not being right.
 
Status
Not open for further replies.
Back
Top