NexusFi: Find Your Edge


Home Menu

 





Tradeable Programmed Price Stream TPPS


Discussion in Cryptocurrency

Updated
    1. trending_up 1,624 views
    2. thumb_up 0 thanks given
    3. group 1 followers
    1. forum 3 posts
    2. attach_file 0 attachments




 
Search this Thread

Tradeable Programmed Price Stream TPPS

  #1 (permalink)
 tpredictor 
North Carolina
 
Experience: Beginner
Platform: NinjaTrader, Tradestation
Trading: es
Posts: 644 since Nov 2011

I like this idea of a tradeable programmed price stream. Let us first define the goal of trading for most retailers: make money by attempting to either predict future price or profit from statistically relationships. This is the purpose of markets for traders. They are a means for making money. It is not the small traders jobs to support the economy or help the exchanges or banks make money from trading fees. Next let us demonstrate that real markets may not be ideal trading vehicles for profit, why?

1. Economist state that markets are designed to be efficient. They are stated to be random walks without memory.
2. Not all trades are equal. A large trader or group of large traders can unduly influence the market in any number of ways such as running stops, seeking speed advantage to the exchange, gaining insider information, etc.
3. Markets can also be manipulated by government powers with disastrous consequences-- for traders. These events are known as black swan events.
4. There are usually more frictions for the small trader. Primary futures markets are designed for large traders. Any sort of secondary market is always biased to favor large traders, as well.

The solution to all of these problems is to recognize that trading is a relative advantage game and to create a pure programmed price stream. As a pure programmed price stream, one can specify the goals or outcome that should be achieved. For example, it could be programmed to produce a certain distribution of profitable traders. A simplest example might a goal of making it such that at least 30% of traders are profitable and at least half of traders don't lose more then a certain amount at a certain -- not sure this is a right goal but this could determined -- but basically instead of rewarding the top 1% of traders a programmed price stream could attempt to return some greater percentage. More over, a programmed price stream could have a self-adjusting difficulty level to make the market less efficient or more efficient to meet its goals. Also, there would be no need for high frequency trading and trades could be limited to some reasonable level such as once every 5 second. Because it is programmed, the source would be available, and no gaming could occur.

A question is how do you get people to trade the wrong side of the market. There are various ways. One solution would be to keep the orders hidden and then after each turn you pay out the winners from a total pool. I'm not too keen on that because I like to read the order flow. But, you could create a synthetic order flow. The importance is that the actual orders wouldn't influence the market meaning that a large trader couldn't move it or have any advantage. Your payouts would just change based on the number of traders that predict it right. For example, say 10 traders are trading the market and all 10 bet long. Well they all get their money returned with no loss. No harm no foul. However, the programming would be designed to dynamically get harder to make it such that only the best win. Now let's say there is some randomness to that and you will have certain cases where 3 out of 10 win or 5 out of 10. The money would be returned in some proportional fashion to the bet size.

But what if a large trader trades the wrong side of the market on purpose to mess up the programming? The programming would be dynamically random on some random sliding time scale. In this way, let's say the market is at an easy level because there aren't many people trading. So the patterns are really easy.. market is range or trending clearly. A large trader comes in and tries to bully the market by losing some money to make it think the market is actually difficult and change its programming -- well in this case they would need to lose money to do that and the change would occur at some random time in the future.

Also, what is really nice is that as a pure programmed price stream, the type of black swan events that plague markets wouldn't exist while also you would always presumably have good volatility for trading. There would be no frictions involved because beyond the initial margin deposit there would be no exchange of money.

In the idealized version, the machine must be run on a distributed set of servers with the source code fully option to inspection and the outcome of every turn verifiable but no future turn computable. The price stream would have to have a memory. If enough people traded the programmed price stream, it is presumable the market would become efficient. However, it would become inefficient as soon as traders started to lose. And the way it become efficient or inefficient would be designed as a goal of the system. More over because there would be no frictions, the efficiency would not be exploited for any entities gain. The price would have a memory by design but the weighting of that memory would obviously have to change. It would be ideal that not just simple state patterns are programmed but complex patterns are also developed. This would allow for any number of predictive patterns to eventually be arbitraged away without having to change the global difficulty level. This would allow for traders to gain specialized advantage.

I'm fairly sure this could work if you trust a central authority but I am not sure it is mathematically or technically possible if you aren't willing to do that. Obviously trusting a CA is not ideal but might be worthwhile to get it up and running. A friction-less token system would work best for payouts. In theory something like Bitcoin would be ideal except Bitcoin is not stable and it is not frictionless and not suited for micro-payments (sigh).

Curious about thoughts on this.. could it work? I do not think this falls under traditional gambling. How would you program in a price stream with a memory but also make the difficulty level always changing? As a trader, the benefit is that if you are better at predicting the market (price stream) or finding patterns you will profit more just like in actual markets. However, you have none of the disadvantages of traditional markets. And unlike futures markets, the market would be zero sum and not negative sum. The only downsides I can think of are (1) only pros might play the game whereas in traditional markets you presumably have some non professionals, (2) you might not get enough trade volume/dollar volume as it might be the converse that only small traders trade, (3) the idea is that we create an optimal trading market for the discretionary trader but it is possible that due to some combination of factors the programmed market becomes super difficult to play/win.

So, obviously I agree if enough trade it then it would be efficient. However, the distribution or effects of the efficiency could be a function of the goal state. Also, I think a really good trader should still win more.

Started this thread Reply With Quote

Can you help answer these questions
from other members on NexusFi?
My NT8 Volume Profile Split by Asian/Euro/Open
NinjaTrader
Exit Strategy
NinjaTrader
The space time continuum and the dynamics of a financial …
Emini and Emicro Index
Deepmoney LLM
Elite Quantitative GenAI/LLM
ZombieSqueeze
Platforms and Indicators
 
Best Threads (Most Thanked)
in the last 7 days on NexusFi
Get funded firms 2023/2024 - Any recommendations or word …
61 thanks
Funded Trader platforms
39 thanks
Battlestations: Show us your trading desks!
26 thanks
NexusFi site changelog and issues/problem reporting
24 thanks
The Program
17 thanks
  #2 (permalink)
 tpredictor 
North Carolina
 
Experience: Beginner
Platform: NinjaTrader, Tradestation
Trading: es
Posts: 644 since Nov 2011

A simpler but similar concept is creating frictionless "programming contract", i.e. defined by programming, based markets:

https://cointelegraph.com/press-releases/spectre-launches-the-alpha-version-of-the-worlds-first-broker-less-trading-platform

--

For the concept of TPPS, I was thinking you could also make it such that all price movements must have be predictable and nothing can be random. All traders will trade against the ML-agent, the master AI, which will seek the goals of the system. The goals of the system will be similar to that of efficient markets except it must work within predictable price streams and could have higher level goals such as making it such that a certain distribution of traders are profitable. So, let's imagine in actual markets only about 5% of traders are profitable and they tend to be the larger traders, the master AI could aim to make up to say 30% of traders profitable.

One way it might achieve this goal would be let's say it introduces a price pattern into the market. And if you pick up on it first, you're trading and you're doing well. However, once enough traders start trading it then the machine must transition to a new set of predictable price patterns. So, the programmed price pattern will never be random. But, the AI might need to adjust the price memory as well dynamically. Of course, ideally the system will be able to create a number of overlapping patterns within different time frames. Obviously, you will need to be better then other traders to win much. However, these sorts of markets/price streams would be designed for winning.

As for pay outs, it might be that you get payed out in skill tokens and have to trade a certain amount before you can collect them. Alternatively, you would need to have a turn based approach in which case you might need to introduce a turn fee, i.e. ante.

Started this thread Reply With Quote
  #3 (permalink)
 tpredictor 
North Carolina
 
Experience: Beginner
Platform: NinjaTrader, Tradestation
Trading: es
Posts: 644 since Nov 2011


It sounds strange but it makes a lot of sense really. Today, most of the trading is algorithmic, and the natural evolution is toward algo generated markets. One useful function of algo generated markets is that large traders have undue influence in traditional markets. In an algo generated market, all traders are equal. And, if it is technically possible to do away with a central authority, you also do away with any possibility for insider trading, cheating, gaming, etc. Of course, I am thinking about it as a way for strong traders (like myself) to profit. But, another type of evolution would be algo-regulated markets where a super AI might set policy.

As for summarizing the benefits, it is basically a market generated for humans or desktop computers to trade. Unlike traditional markets which seek to optimize efficiency and in which exchange fees and regulation disadvantage all traders: the algo generated frictionless market seeks to make it such that some percentage of traders can make a living or at least be profitable trading against other traders. It might be such that all the traders in the game own the exchange as they would help to run the computers that compute the game.

Right now, I only see a way to do this based on traditional markets like running a binary options via a token based programming-contract (Etherium) like model or a central authority. I do not like the idea of a CA because you have to trust the CA. One way around a CA while using a CA might be that everyone in the "game" gets to compute something and keep it secret while another piece is returned that ensures the integrity of the game.

In other words, a CA runs the game engine. However, each trader provides a piece of data that is important to the game and that they keep secret but that they can't change. The way it works is imagine a series of turns UP DOWN DOWN UP. This is computed on local machine. A signature is shared at beginning of the game ensuring the local trader can't change the sequence. However, the signature doesn't reveal the sequence. Every trader computes such a signature.

Started this thread Reply With Quote
  #4 (permalink)
 tpredictor 
North Carolina
 
Experience: Beginner
Platform: NinjaTrader, Tradestation
Trading: es
Posts: 644 since Nov 2011

Some of the notes in the Spectre paper are interesting. (Disclaimer: I have did zero research on the company/group behind this project nor have did any research into legalities, potential for scam, or anything else. As an aside, these are similar ideas to some of the simpler ideas I shared but no really equivalent as part of my interest is truly frictionless trading environments).

https://www.spectre.ai/media/spectre_whitepaper.pdf?ver=1.8

The average win rate of inexperienced traders is around 35-50%.6 Experienced traders,
however, who spend time in learning technical or fundamental analysis in order to gain
an edge and be able to beat the market, can at times obtain a 60-70% win rate, although
emotions and greed tend to get the better of them, thus suppressing their win-rates
below 60%, over time.7 Since SPECTRE’s liquidity pool pays out 75% on winning trades
(and as high as 93% in certain conditions) but keeps 96% of all losses, traders on average
need to maintain at-least a 57% win-rate in order to break-even. This is difficult for the
masses to achieve and therefore, average win rates in the system are likely to hover
under 48-53%, resulting in perpetual growth in SPECTRE’S liquidity pool and
contractually, the size of special dividend payment to token holders (Exhibits C & F)

---
Not evaluating the payout to token holders, the ideas expressed on win rates seem reasonable. It is difficult for most technical analysis methods to achieve win rate over 55%.

Started this thread Reply With Quote




Last Updated on September 15, 2017


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts