{quote} Oh, but ChatGPT assumes that you are providing the tick data... You can upload that to Perplexity too, and ask Perplexity Labs to create an interactive backtesting app for you, for example.
- #662
- Edited 8:38am Nov 17, 2025 7:13am | Edited 8:38am
{quote} Oh, but ChatGPT assumes that you are providing the tick data... You can upload that to Perplexity too, and ask Perplexity Labs to create an interactive backtesting app for you, for example.
- NQ 1-Minute Chart
- 1% of current equity per order
- Random BUY/SELL entry every 30 minutes during RTH
- Stop Loss = -10 points
- Trailing Take Profit = +30 points with a 10-point trailing step
- Max. 3 consecutive daily losses
- Max. -3% Daily Drawdown
- Max. -6% Weekly Drawdown
{quote} Thank you for sharing your results. Honestly, I expected those results, as an edge purely from the risk/reward ratio is unlikely. Maybe there is a small edge to the long side, as the skew of the underlying is upwards over the years, but will hardly overcome the trading costs. Grüße aus dem Rheinland
I'm not yet sure if the EA I programmed followed the described rules as precisely as outlined in the Perplexity PDF (since some of them are somewhat vaguely defined). I tried to recreate it as best I could. I'd like to check this more thoroughly. Unfortunately, I don't currently know exactly what instructions Medici gave Perplexity to produce such a nice strategy description with all the rules, performance metrics, graphs and so on. I've done some reading, but I haven't been able to replicate the exact AI prompt yet.
Medici said they can be good "pointers," but honestly, if I replicate such a strategy exactly and don't even come close to the performance figures that Perplexity spits out, I wonder how trustworthy the AI is when it comes to strategy development and backtesting. I don't know if it's due to the data the AI used for testing (or whether it actually used real historical data for an accurate backtest the results are simply fabricated). Or maybe eben the described strategy rules in that PDF are completely different from what I've tested so far.
Meanwhile, my ORB system has achieved a +17% return so far under -5.3% Max DD, and that's not solely attributable to Gold. The DAX seems to be getting closer to the same profitability. Nastaq is disappointing.
Gold: +7.8%
GER30: +7.3%
NAS100: +1.0%
I still have no idea if or how I'll achieve my goals of finally being able to live off this someday. Not only do I seem to lack a good strategy, but also the necessary capital, and I can't seem to find any effective help on how to proceed without "wasting" more years of my life.
{quote} I think, that's the point, but nevetheless and out of curiosity, I will set up this stragegy in my EA framework. The we have a second set of results, that either confirm you findings or reveal a discrepancy in the rules.
I just asked Perplexity the following question:
Do a comprehensive backtest evaluating a random-entry trading strategy
The answer was:
A comprehensive backtest of a random-entry trading strategy over one trading year shows the following results: The strategy randomly enters long positions with a 10% daily chance, holding each trade for up to 5 days or until a 2% stop loss is hit. Despite the randomness, this approach yielded a surprisingly high cumulative return and Sharpe ratio in the simulated scenario. If you want, the strategy can be further analyzed or optimized with different parameters or compared to benchmark strategies. Let me know if you need additional details or visualizations....
In which market was this backtest conducted?
The answer was:
The backtest is based on synthetically generated price data that does not represent any specific real-world market or financial instrument. The price data was generated using a simulated daily return pattern to demonstrate the functionality of the random entry strategy. Therefore, it is a purely hypothetical simulation with no reference to an actual market. If desired, the backtest can be performed using historical data for a specific market or instrument.
This likely explains the overwhelmingly positive results mentioned in the original PDF "Optimal Trade Management and Money Management Strategy.pdf," where Perplexity cites a "31.547% return with only 4.4% maximum drawdown."
Apparently, Perplexity had a rather unrealistic summer night's dream
- #667
- Nov 18, 2025 9:38am Nov 18, 2025 9:38am
- Joined Jan 2016 | Status: Trader | 1,006 Posts
{quote} Thanks a lot. Yes, I wanted to share my EA on it too, but if it has absolutely no positive expected value, I don't want to waste people's time with it. I just asked Perplexity the following question: Do a comprehensive backtest evaluating a random-entry trading strategy The answer was: {quote} In which market was this backtest conducted? The answer was: {quote} This likely explains the overwhelmingly positive results mentioned in the original PDF "Optimal Trade Management and Money Management Strategy.pdf," where Perplexity cites...
Trading is a game of impermanence. Edges appear, get discovered, and disappear — often faster than we expect.
Full article from Raen :
Edges aren’t meant to last
In trading, everyone is chasing the same thing: an edge. That small, consistent advantage that lets you make money. But here’s the uncomfortable truth — no trading edge lasts forever.
A note worth reading
Our CEO Ryan Wright recently released a Substack on this entitled:
Edge isn't yours: what actually works in modern markets
It is worth reading. But as I ever search for topics to write on, I thought I’d add my experience in this having seen it in practice for over 25 years. Some good old war stories as it were!
The pattern we spotted
Going back to 2008 (ish) I had two young traders on my desk, one of whom went on to be fairly public, but at this point were at the start of their careers. They noticed a pattern that occurred around the close of European indices. The S&P, mostly collapsing as we started to enter the Financial Crisis period, seemed to begin to have no influence on the DAX. The DAX just refused to follow suit and in some instances went up.
Trading it in real time
Then the next morning, having been propped up all night, the DAX would adjust and open sharply lower, catching up to the S&P on the open. They spotted this for a few days then asked if they could trade it, holding overnight.
Size fast, bank it
It went on to be later found out that a rogue trader called Jérôme Kerviel had been so long in the DAX, unauthorised, that to afford the margin requirements and obscure his P&L he was artificially holding the DAX higher.
However, that's not the point. The point is that over a week these two traders spotted it and then began trading it. They didn’t worry about how long it would last. They just started to trade.
They then ramped up their size very, very quickly.
I would say it was a major turning point in their careers. They went from small five-figure accounts to very large six (close to seven)-figure accounts.
When the music stops
And then it stopped. I can’t exactly recall but it may have only lasted a couple of weeks, certainly no more than three. Kerviel got found out. SocGen management unwound the positions and the edge disappeared.
The lesson
My traders’ profits didn’t. They moved on, using their newly-found wealthy trading accounts to capitalize on the remaining period of the Financial Crisis. And that is the point. Edges do come and go. I’ve seen so many of them drift in and out over the years.
Yet the consistent point for all of the traders who “make hay when the sun is shining” is that they don’t spend time worrying about the edge lasting or what they may do if it disappears. They run with it, bank the profits, and make peace with the nature of the game.
Edges can disappear for a number of reasons
Markets adapt faster than ever. They evolve.
As soon as a strategy starts showing reliable profits, it attracts attention. Liquidity floods in, spreads tighten, and the inefficiency that gave you an edge vanishes. Think of it like a gold rush: the first few miners strike it rich, but soon the crowd arrives, and all that’s left is dirt.
Data and technology flatten the playing field.
The democratization of technology makes it harder to stay unique — and edges born from superior tech tend to decay quickly.
Behavioral edges fade as traders get smarter.
Once enough traders recognize a behavioral bias (say, overreaction after earnings), they start trading against it. As more people act on the same insight, the edge self-destructs.
The “feedback loop” problem.
Success can destroy your own edge. When a strategy performs well, traders often scale up — increasing position sizes, running the same signals across more assets, or publicizing results to attract capital. That attention leads to crowding. In short, your own success can be the seed of your downfall.
Structural changes reset the game.
Regulations, macro shifts, and new market structures constantly reshape the landscape. A strategy that worked in the low-volatility era of 2015–2019 might collapse in a regime of higher inflation and interest rates. Edges are often regime-dependent, thriving only under certain conditions — and dying when those conditions change.
How Do You Stay Ahead?
- Keep evolving. Treat every strategy as temporary. Continuously research, test, and adapt. Don’t overly worry about what may or may not be in the future. Focus on the now. The one good trade, right here, right now.
- Accept the game you chose to play. Acceptance is the beginning. Once you can accept what is, you can stop trying to change the unchangeable. It will give you the freedom to take advantage of the opportunity in front of you in its entirety. With that sort of single-minded focus you can avoid the pitfall of wishing you had taken advantage of the situation in hindsight — and in turn avoid the loop of repeating the same mistake when the next opportunity comes knocking.
- Stay humble. Assume your edge is decaying the moment you find it. That mindset keeps you sharp.
Final thoughts
Trading is a game of impermanence. Edges appear, get discovered, and disappear — often faster than we expect.
The traders who survive aren’t the ones who found a “holy grail” system; they’re the ones who accept that edges fade and build a process, often starting within, that evolves faster than the market itself.
I might be wrong , I have been wrong before
- #670
- Nov 18, 2025 1:25pm Nov 18, 2025 1:25pm
- Joined Jan 2016 | Status: Trader | 1,006 Posts
I might be wrong , I have been wrong before
- #674
- Nov 19, 2025 9:35am Nov 19, 2025 9:35am
- Joined Jan 2016 | Status: Trader | 1,006 Posts
I might be wrong , I have been wrong before
{quote} So be it, but that expectation has implications and looking at three months of trading data then poses a statistical question: How likely is this trading system to deliver on your expectation? You would like to know the answer to that question, I believe? If so, then some sort of statistical analysis would be needed. The ratios (profit factor, Sharpe ratio, etc) give an indication, but that's all. So how will you know the answer to the question? Only by some calculation. In my experience, Monte Carlo Simulation beats the ratios by far, in...
NinjaTrader is working now. I have now purchased the CME Bundle (Level 1) - Top of Book from NinjaTrader for $12/month and thus at least now have a demo account that does not expire and has real-time market data. They say you need a $50 daily margin to trade MES? Meaning if I deposit $50, that should be enough to open a position with the smallest possible contact size during regular market hours, right?
