Confidence Score: The Ultimate Strength Meter
This is the last part of the "getting started" series for The Stock Archeologist. We've talked about what signals the program looks for. Now, let's talk about how much conviction the program assigns to those signals. That's the Confidence Score.
It's a number, 0 to 100. Higher means the program sees more supporting factors aligned with the primary signal trigger. It’s a weighted average of several different elements, designed to give you a quick gauge of the overall strength of that particular setup, beyond just the basic signal criteria. Think of it as a multi-factor verification layer.
The score isn't pulled out of thin air. It synthesizes data points that provide context and potential confirmation (or contradiction) to the main signal. Here’s the gist of what influences it, roughly in order of impact:
Projection Quality (Heavy Weight): This looks at the historical backtests for the specific signal type on that stock. How consistently did similar past signals lead to positive outcomes? Is the projected range (max vs. min potential gain) tight and skewed positive? Is the projection based on a decent number of historical examples (test count)? Strong historical performance boosts confidence significantly. Weak or sparse history lowers it.
Market Defiance (Heavy Weight): Is the stock moving with or against the broader market (SPY)? A stock signalling a buy while the market tanks shows relative strength – that gets a confidence boost. Conversely, a buy signal when the whole market is roaring might be less impressive. This uses beta (a measure of stock volatility relative to the market) for a smarter comparison.
Market Context (High Weight): Related to defiance, but simpler. Is the overall market trend (short-term and long-term, based on SPY) bullish, bearish, or neutral? A buy signal in a strong bull market gets a higher score than one in a bear market. Obvious, right? The score penalizes signals fighting a strong opposing market tide.
Earnings Timing (High Weight): Is an earnings report just around the corner (or did one just happen)? Earnings are volatility events, often unpredictable. Signals occurring too close to earnings (upcoming or recent) get their confidence reduced. We prefer clean setups away from earnings noise.
Analyst Consensus (Medium Weight): What are the professional analysts saying? The score looks at the consensus price target (is the current price far below the target? Good.) and the overall analyst rating distribution (Buy/Hold/Sell). It’s a secondary check, adds some texture.
Fundamentals (My Caesar Score) (Medium-Low Weight): Is the underlying company solid? My proprietary Caesar Score (you know, named after me) gives a quick read on financial health. A fundamentally sound company adds a bit more confidence to a technical signal.
Seasonality (Low Weight): Does the stock historically perform well during the current (and potentially next, if the projection spills over) month? It’s a minor factor, more context than driver, but included for completeness.
Because not all signals are created equal. A stock might trigger a buy signal, but if the market's collapsing, earnings are tomorrow, and historical tests look shaky, you need to know that. The Confidence Score distills that complex context into one number.
It helps you:
Filter: Quickly see which signals have more supporting evidence.
Prioritize: Focus your limited research time on the higher-conviction setups.
Manage Risk: A lower score might warrant a smaller position size or tighter stop-loss, if you decide to trade it at all.
Think of it as a probability gauge, not a guarantee.
Very High (80-100): Strong alignment. Multiple factors support the signal. Worthy of serious attention.
High (60-79): Good alignment. Most factors are positive. Solid setup, standard checks apply.
Medium (40-59): Mixed signals. Some factors support, others don't. Needs careful scrutiny. Understand why it's not higher.
Low (20-39): Weak alignment. Significant conflicting factors. Tread very carefully, likely needs strong independent justification.
Very Low (<20): Major red flags. Usually best to ignore unless you have a very compelling, contrary reason.
The Confidence Score is a critical tool for interpreting the daily feed. It adds nuance, context, and a layer of risk assessment. Use it alongside the raw signal and the projection chart. Don't trade blindly based on any single number, but use the Confidence Score to rapidly assess the landscape.
That’s it for the introduction. You now have the essential framework for understanding The Stock Archeologist's output. The rest is up to you – execution, research, risk management.
Time to get back to the real work.
Caesar
Great explanation of the scoring, thank you!