6 months TopStep 50K Net Profits

[ironchef]

Current ES price is about 3000.
Let’s say 95% of daily deviation < 5%
It’s 150 points or 7500$ risk per day with 1 contract.

Risk of ruin is 5% with 7.5k margin per contract.
For intraday, 5% deviation is min requirement IMO.

This is last 5 years returns for ^SPX
View attachment 221214

What other conclusions can you draw from these histograms?

 

[Sekiyo]

What other conclusions can you draw from these histograms?

That Daily (1) and Monthly (20) are gamble.
Mostly 50/50. with mean 0.

So price are reverting monthly on average,
From weeklies.

That it takes 2 to 5 days to express a trend.
Maybe retracements happen over a week (-2%)

That above 50 days price is likelier to increase.
Mean 1%(50), 3%(100), 12%(250)
But numbers are biased.

250 days from 5 years isn’t much sample.

...

What do you see ?

 

[Sekiyo]

That Daily (1) and Monthly (20) are gamble.
Mostly 50/50. with mean 0.

That it takes 2 to 5 days to express a trend.
This where we see the most asymmetries.

That above 50 days price is likelier to increase.
Mean 1%, 3%, 12% but numbers are biased.
250 days from 5 years isn’t much sample.

...

What do you see ?

Also price isn’t moving much most of the time.
As 90% of the time, daily return is less than 1.5%

Volatility is clustered.
Come in bulk and goes away.

 

[Sekiyo]

“Jeffrey Ma was one of the leaders of a notorious team of blackjack players from the Massachusetts Institute of Technology. To make money, the team counted cards. Their system had two crucial components. First, team members fanned out and counted cards at a number of different tables in order to determine which tables were attractive. In this initial phase, the players stuck to small stakes. They were playing solely to determine if the cards that remained in the shoe had a relatively large number of high cards. The more high cards, the greater the chance that the player will win a hand. When a player found an attractive table, a teammate would join him and place large bets in order to win as much money as possible. As described in Ben Mezrich's best-selling book, Bringing Down the House, the team could express the attractiveness of the table and how large the bets should be with mathematical precision. Ma and his team were acutely aware of the influence that luck could have and therefore stayed focused on their decision-making process. Indeed, Ma recounts an instance when he lost $100,000 in just two rounds over the course of ten minutes, even though he played his cards just right: “The quality of the decision can be evaluated by the logic and information I used in arriving at my decision. Over time, if one makes good, quality decisions, one will generally receive better outcomes, but it takes a large sample set to prove this.” In other words, he has to place a lot of bets in order to win, because this game involves a lot of skill but it also involves a lot of luck.”

Excerpt From
The Success Equation: Untangling Skill and Luck in Business, Sports, and Investing
Michael J. Mauboussin
This material may be protected by copyright.

 

[Sekiyo]

If I understand it well,
The Pro account is the way to go ?
It's simulated (No fees) ...

Same for profit but you can exercise it anytime.
Get the funded account.

Why would someone go with the funded account ?
Someone that want to withdraw less than 5k ?

 

[TommyR]

imagine if due to the pareto (not normal) almost all data is pointless in finance except in the domain of elections where its very important

 

[Sekiyo]

imagine if due to the pareto (not normal) almost all data is pointless in finance except in the domain of elections where its very important

Statistics and probability is useless without the right distribution.
Like wandering through Paris with the map of New York.
You either throw the map away or buy a ticket to NY.

This is why most of the theory is useless as it assumes a normal (log) distribution.
But we can and do model financial markets with the student T distribution and other tools.

It's said statistics should be used to discard evidence.
Refutation vs Positivism

 

[Sekiyo]

luck={-2,-1,1,2}; skilled={-1,1,1}; unskilled={-1,-1,1};
Sum the addition between random picks from both luck and skilled or unskilled.

Over the short term. This can happen.

 

[Sekiyo]

One way to achieve convexity from a linear payoff is to bet percentages.
Which is to be convex over time rather than through space.

Here you bet 1$ to earn 1.1

Here you bet 1% to earn 1.1%

First I thought about averaging into a position.
But in the end it's just a displacement.

 

[Sekiyo]

I believe that a trader’s job is to :
1. Estimate Risk (Likelihood & Impact)
2. Estimate Reward (Likelihood & Impact)
3. Figure out if the bet is a sucker’s one
4. Size accordingly

So you need models to :
1. Estimate
2. Compute
3. Allocate

If you can’t get an expectancy,
And a Kelly betting size,
Then STOP.

If your models are flawed,
Then you don’t have an edge.
But at least you need consistency.

Most of the time we mistake the map for the territory.
We believe our tools and understanding are the reality.

We should take the trade iif even the worst case looks good.

“Depending on how good, or how bad, something might be compared to my most probable estimate.”

To bet only on favorable situations,
Is to not bet on unfavorable ones.
If downside isn’t cheap then,
DO NOT BET.

I am still waiting for TST to activate my Pro account.