Welcome to the FVEr Invest Platform.
At FVEr, we do data driven investment research to help you make informed decisions about your investments. We use 20 years of historical stock market data to construct fair value estimators (FVE) for various segments of the stock market, using an exponential regression based algorithm.
The Two Pieces of FVEr
FVEr Algorithm and Platform
The FVEr platform is our custom-built algorithm that analyzes market data to identify when major stock indexes are undervalued, overvalued, or fairly priced. It’s the engine that powers everything we do, using proven math and logic to cut through the noise and focus on what really matters: value. Designed for consistency and adaptability, this platform helps take emotion out of investing.
FVEr Trading Strategy
Our proprietary trading strategies are how we put the platform to work. When the data shows markets are undervalued, we lean into risk using leveraged ETFs. When things look overvalued, we get defensive with inverse ETFs. And when markets are fairly priced, we stay steady with traditional index funds. It’s a simple, rules-based approach that aims to keep you positioned for long-term success — without the guesswork.
Who We Are
Trubee Davison
Co-Founder
Trubee is co-founder at FVEr Invest. Among other things, he oversees investment algorithm design. He attended Williams College for his undergraduate degree, and holds a Mathematics PhD from the University of Colorado Boulder. The genesis for FVEr Invest came to him while teaching a class on statistical regression. He resides in Boulder, CO.
George Ferrell
Co-Founder
George is co-founder at FVEr Invest. He oversees investment strategy, and social media. He is a graduate of the University of Houston with a BBA in Finance. During his time at university, he focused on strategy design and personal finance. Currently, he holds the Series 7 and Series 66 certifications. He resides in Austin, TX.
FAQ
-
We use 20 years of historical weekly price data for a particular security, and apply an exponential regression based algorithm that produces a fair value estimator for that security, updated weekly. The algorithm employs an ensemble method adapted from machine learning to robustly mitigate for starting point bias, while also disregarding outlier data via a filtering strategy.
-
We’ve back tested all of our models for many years. When deciding if the model performs well we look for an average percent difference between the index price and the model to be near zero over a 10 year time period, a bell shaped distribution of these differences, and an effective trading strategy which is empirically validates the algorithm.
-
When FVEr residuals suggest an index is undervalued (historically 40% of the time), the strategy allocates to a leveraged ETF—either 2x or 3x—depending on the investor's risk tolerance.
When residuals show an index is extremely overvalued (historically 10% of the time), the strategy allocates to the inverse ETF, allowing for profit potential during downturns.
A neutrally valued residual (historically 50% of the time) will prompt the strategy to allocate to the unleveraged index.
-
Not at this time, but we may launch that in the future.
-
Yes. We cover a few selective international markets. Due to the higher volatility of these markets, we have less conviction in the model applied to the international setting.
-
We’ve found that 20 years is the optimal time frame for running our model — it captures enough market cycles to be meaningful without introducing noise from outdated market dynamics.
-
We could track Bitcoin — technically, it’s not difficult. But FVEr is built to focus on long-term, data-driven investing, and we aim to offer strategies that people can feel confident in for decades. Right now, Bitcoin still behaves like a speculative asset, and the historical data available simply isn’t deep enough to meet our modeling standards. Until it matures into a more stable and predictable part of the market, we’re choosing to stay focused on assets with proven long-term fundamentals.