EyeQuant is an AI (artificial intelligence) tool that systematically analyzes the market to provide market signals for stocks, ETFs, cryptocurrencies, and other assets.

It aims to save you much of the work of finding and analyzing information yourself. Instead of reviewing hundreds of charts or assets, EyeQuant identifies those for which its models detect a trend and presents them as signals.

These signals come from quantitative models evaluated through out-of-sample backtesting, using market periods the model did not previously use to define its parameters.

EyeQuant systematically monitors a broad universe of assets and analyzes their behavior using quantitative models.

When an asset meets the model's conditions for the start of an upward trend, EyeQuant generates a signal shown as a New Idea. While those trend conditions hold, it is shown as Trending. When they no longer hold, EyeQuant generates an end signal shown as Ended Today. It is then shown as Inactive, meaning EyeQuant AI currently has no active upward-trend signal for that asset.

The process is systematic: signals follow model rules, not a person's opinions, news, or discretionary decisions.

Signals are information generated by EyeQuant models. They are not recommendations to buy, sell, or hold, nor assessments of whether an investment is suitable for you.

A trend is a relatively persistent movement in an asset's price in a given direction.

Markets rarely move in a straight line. An asset may rise over several weeks and still fall on some days. EyeQuant looks beyond each individual move to assess whether the overall price behavior is consistent with a trend.

When EyeQuant shows an asset as Trending, the model has detected the conditions it uses to identify the start and continuation of that kind of movement.

This does not mean the price will rise every day. It means that, under the model's rules, there is currently a trend signal.

A quantitative model is built and adjusted using a training period: historical data used to define its parameters.

Out-of-sample evaluation then tests the model on different periods that were not used during training. The model generates signals from data it did not previously use to tune its parameters.

This form of backtesting seeks to reduce overfitting and provide a more demanding assessment of how the model might have performed on information it had not seen while being built.

An EyeQuant signal is information generated by a quantitative model. It is not a prediction, a guarantee of returns, a recommendation, or investment advice.

When a signal appears, the asset currently meets the conditions the model uses to identify a trend. You can review the signal and its background and, as a good practice, do your own analysis or discuss it with your investment adviser before making a decision.

EyeQuant signals are a starting point for your analysis, not recommendations or investment advice.

Not every signal will succeed. A strategy can have positive results over time even if some individual trades end in losses.

EyeQuant aims to provide objective, systematic information to complement your own analysis. The final investment decision always belongs to you.

The EyeQuant AI Watchlist lets you select the stocks, ETFs, cryptocurrencies, currencies, or other assets you actually want to follow.

Instead of reviewing dozens of ideas available in EyeQuant, you can make a selection based on your interests, knowledge of particular assets, or any other criteria relevant to you.

The Watchlist acts as a personalized focus: it helps you follow the assets you care about and quickly see which are New Ideas, which are Trending, which Ended Today, and which are Inactive according to EyeQuant AI.

We start with an initial value of 100 and simulate how the ideas in your Watchlist evolve over time. The calculation updates when an idea enters, closes, or reaches the end of the selected period.

When an idea closes, its value becomes available for new opportunities. If that value is not enough to add a new idea, only earlier positions are reduced, without artificially increasing any of them.

This reflects the actual sequence of ideas without assuming they always have the same weight or are rebalanced every day. If a relevant date falls on a day when the market is closed, we use the most recent available closing price to continue the calculation.

ABS Gate Analyzes recent signal patterns that have historically produced better results. It then applies a fixed filter to show only the opportunities with the highest confidence under that analysis.

STRICT ALLPRE Uses a broader history to identify risk-and-reward combinations associated with better results. It applies a stable, more demanding criterion to prioritize signals with greater historical consistency.

RATIOS T10 Evaluates each signal's potential reward relative to its risk. Rather than using a fixed threshold, it compares the signal with recent opportunities and shows approximately the best-positioned ones in the Top 10.

RATIO T10 Similar to RATIOS T10, but adds a check of the quality and consistency of risk data before selecting the best-positioned signals.

M05 Combines several sources of information: the asset's context, the market, the risk-reward relationship, and the characteristics of the trade. It seeks opportunities that meet favorable conditions under more than one criterion.

M10 A more comprehensive version of M05. In addition to assessing the trade and its risk-reward relationship, it places more weight on the asset's and market's recent behavior.

Each model uses different criteria to prioritize signals. Being selected by a model does not guarantee future results.