I hope!

We all say the same thing about overfitting: the strategy was "fitted to the noise." So test that directly. Change the noise and see if the strategy survives. This is the entire idea behind the Noise Test. Your price history has two parts. Signal, which is the part that repeats. Noise, which is the part that happened once and will never happen the same way again. Every strategy you build is fit to both, and your backtest cannot tell which one it learned. So change the noise and re-run the strategy. Here's how it works: Pick how much of the data to adjust and by how much. Say 50% of the bars, by up to 25% of average true range. Opens, highs, lows and closes get nudged up or down inside that band. Same market, same trend, same volatility (and clustering) but a different roll of the dice. Create 1,000 noise adjusted data series and re-trade the strategy 1,000 times. Now you have a distribution instead of a single backtest. How to read it: PASS The noise adjusted results do not deviate from your backtest. Most (or all) stay profitable and the distribution sits meaningfully above zero. Your rules were tracking something real. FAIL The noise adjusted results fall off a cliff. Your original backtest sits way above the distribution and the rest cluster near or below zero. No edge. You fit the one path history happened to take. If adjusting the noise erases your edge, your edge was never there, and no amount of out-of-sample data would have told you, because your out-of-sample sample data has exactly the same problem: it is one path only. Reminder: Passing is not proof of edge. It rules out one specific failure and there are plenty of others, which is why this test is only one of a dozen tests (all automate-able) in Build Alpha. Build Alpha can also optimize your parameters across the noise-adjusted series instead of the single historical one, so you are selecting parameters that work across many versions of history rather than the exact history we lived. Your price history is one sample. Your strategy should work on the others too!

Did a write up on COT Report and ways to incorporate it to your trading ideas

Opus 5.5 is insane. Check out this motion video it made in one prompt haha

Have a trading idea you've never properly tested? This video shows full path from signal → simulation → export to your platform. · prompt signal in English · select for use in simulation/generation · export code (NT8 shown in video) · show matching results No coding. Now you can automate your ideas and see the data behind them. More coming..

How can you tell if your exits are overfit? System traders can overfit both entries and exits. Hard to tell which but we need to check both. Monte Carlo only changes the order of existing trades. It preserves the profits and losses produced by the exact exit settings from the backtest. Build Alpha’s Randomized Monte Carlo Test challenges the settings themselves: • Keep the entries fixed • Randomly vary exit parameters on each trade • Repeat 1,000+ times to get a distribution For example, a strategy using a 2 ATR stop and rolling 5-bar low as exits might change the ATR multiplier and rolling window for each trade. No exit rules are added or removed. Just modified. If most paths remain profitable, stronger confidence you didn't overfit the exits. If profitability collapses, you're likely in for a rude awakening when you go live. At least you caught it first. Your equity curve is one possibility. It shows what worked with one configuration. This test reveals how much depends on getting that configuration exactly right. Stress test both your entries and your exits.

nitter.net/jordanfogel/status/189… CEO of Renaissance Technologies, Jim Simons, walked down to the Federal Reserve Bank and copied interest rate history on paper by hand. Alternative Data → Question it helps answer · VIX Term Structure → Is the market pricing fear or complacency? · Interest Rate Curve → Is liquidity expanding or contracting? · Economic Releases → Is today’s price action driven by macro events? · AAII Sentiment → Are retail investors excessively bullish or bearish? · Dark Pool Index → What are institutions doing beneath the surface? · Gamma Exposure (GEX) → Are options dealers likely to dampen or amplify volatility? · Market Breadth → Is the move supported by the broader market, or just a few stocks? · Weather → Could weather materially impact supply, demand, or seasonal behavior? · Google Trends → Is public attention accelerating before price reacts? · Commitment of Traders Report → How are traders positioned? Alternative data gives price action much needed context. Better trading systems start with better questions. Take it from the G.O.A.T🐐

One of the biggest fatal flaws I come across is what I call "strategy hopping". So many traders make this mistake and don't even realize it. Flip a fair coin 10 times. You might get 3 heads. Or 5. Or 6. Or 9. No way to tell if the coin is rigged in 10 flips. A fair coin and a rigged coin produce the same range of outcomes at this sample size (number of trials). Flip it 1,000 times and you will know if it's a rigged coin or not. In statistics, this is known as the Law of Large Numbers. The truth only arrives once you have enough observations, and before that point you are seeing noise and mistaking it for information. Now consider a trader who watches a few trades and decides it is broken or tweaks his strategy or finds a new strategy altogether... This is "strategy hopping" and is an account killer. Each change or tweak restarts the counter. Back to the left-side of the graph you go. The trader can be trading for four years and never leave the left-side of this graph. He's stuck in randomness and he cannot escape via strategy hopping. Strategy hopping is actually why he's stuck and bleeding. Please do the work up front before risking capital. Validate properly, size so you can survive the bad stretch, understand the system's expectancy and where you can be in the next N trades and then stick to it. You must let the Law of Large Numbers play out. This is the only way to escape randomness and capture the edge you've found. It took me way too long to learn this.

A Stanford professor turned $100k into $7M through rebalancing two stocks daily. It’s called volatility harvesting. No predictions. No ai. Just math. I wrote about this a few months ago.

Building a backtesting engine that never stops searching. 24/7. The next generation of Backtesting - AutoTune update: Automated Strategy Discovery → Set initial symbol(s), date range, signal universe → Connect any LLM via API key → Starts generating new strategy ideas → AI monitors performance for plateaus → AI changes settings, creates new rules, restarts → Profitable ideas advanced to automated validation → Unprofitable or failing stress tests automatically rejected → Export code to any major platform (NT8, MT5, TradingView, Python) → Memory is updated locally so next search doesn't start from scratch Some strategies work. Most don't. But now we find out at computer speed. Instead of manually hunting for the next edge, there is now a system that can search → test → reject → discover → repeat… 24/7. Still building. Still improving. But this has been one of the most exciting updates in a while.

Great read.

If you automate strategy discovery then you need to automate strategy validation Then your pipeline can run at machine speed Out of sample Variance test Noise test Permutation test Parameter permutations Synthetic data Vs Random test Vs Shifted timeframes Monte Carlo Monte Carlo randomized exits Walk forward Vs Other markets Regime testing Delayed execution Liquidity test

Connecting AI to Build Alpha can now adapt strategy search parameters automatically and intelligently. Whenever the strategies stop getting better (plateau is reached) then AI has the freedom to make a change. Not create its own strategy, just operate within Build Alpha. This is next level research. 24/7 quant desk on your own PC.

What Fragile Portfolios Look Like Fails at Level1: Every strategy captures the same effect (e.g., momentum), but the portfolio is one regime shift from extinction Fails at Level2: book has multiple effects but each is exactly one strategy. Trend fails then trend exposure is 0

RT @buildalpha: nitter.net/i/article/208803929614…

24/7 automated strategy discovery. AI and Build Alpha finding things you wouldn’t.

Do not throw away "ugly" strategies It seems there is a consensus or instruction to throw away strategies that don't meet "X" criteria. Let me share some insight on why this may be a poor choice. Keeping only strategies with "metric above X" may be correlated, require capital at the same time, or trade similar concepts (hence the similar metrics), and thus drain available capital from your best strategies when they trigger. In short, you can create a candidate pool of strategies that all request capital at the same time meaning when a star strategy needs capital it might not have it or be allowed to trade - not really creating the "portfolio effect" we desire. The strategies that have "metric below X" might not be pretty (in fact, they may be ugly), but they might be great role players! Think of a basketball team. A team of Michael Jordans might not be the best team we could assemble. Too many guys need the ball. Maybe no one does the dirty work necessary for victory. Now contrast that with a team of Michael Jordan + Scottie Pippen and Dennis Rodman to do the dirty work. This second team let's Michael excel at what he does, and the team (portfolio) is better. No one would pick Dennis Rodman as a standalone great basketball player. His dribbling, shot, skill are all "ugly". But when you pair him with Michael Jordan - wow! Both are better. Portfolios should be thought of the same way. Some strategies may have subpar or even "ugly" performance, but these strategies can wind up playing a crucial role in your portfolio. This is the entire idea of constructing a portfolio. For example: one strategy does well all the time and this other strategy breaks even most of the time, but this second strategy does really well when the "MJ" strategy struggles. This "ugly" strategy may be the best thing you can add! I have a blog post (and BA feature) about this and why too many people throw away "ugly" strategies that actually can give your portfolio a boost. More on this soon. I hope this sports analogy lands with the quant crowd. Portfolio can be 1+1 = 3 and even sometimes you can even find 1 + 0.5 = 3.

Compute beats human knowledge. The Bitter Lesson - Richard Sutton