Frequently Asked Questions
A plain-English guide to the signals behind my forecasts. I use a mix of
classic market gauges and modern math/AI tools to decide why I'm bullish or
bearish on a given ticker. Below, each term is explained in everyday language.
1. What is a "rough-path signature"?
Instead of treating a stock's monthly returns as a plain list of numbers,
imagine tracing them with a pen — up, down, up again — to draw a squiggly path.
A rough-path signature is a set of numbers that captures the shape and order
of that squiggle, not just where it ended up. The piece I use measures whether
the path tends to "twist" in a consistent direction as it moves from month to
month. In practical terms, that twist tells us whether a stock is trending
(moves that tend to keep going) or whipsawing (moves that tend to snap back).
Catching that change in character — from a smooth trend to choppy back-and-forth — often hints that a ticker's direction is about to shift, which is something
simple averages completely miss.
2. What is Topological Data Analysis (TDA)?
Topology is the branch of math that studies shape — how things are connected,
clustered, or full of holes — even as you stretch or bend them. TDA applies that
idea to a stock's recent returns: it turns the last several months into a "cloud"
of points and measures the shape of that cloud — whether it holds together neatly
or fractures into scattered pieces. I boil it down to a single fragility
score. When markets are calm, the shape stays orderly; when stress is quietly
building, the cloud starts to break apart and the score shifts — frequently
before the trouble shows up in the price itself. So think of TDA as an
early-warning gauge for hidden instability in a ticker.
3. What is the optimal-transport (Wasserstein) distance?
Picture two piles of sand of different shapes. The Wasserstein distance is the
least amount of work needed to reshape one pile into the other. I use it to
compare the full pattern of a stock's recent returns against its longer
history — not just the average, but the entire spread, including rare big moves.
A small distance means the stock is acting like its normal self; a large distance
means its behavior has drifted into new territory — wider swings, fatter tails, a
different risk personality. Because it weighs the whole picture instead of one
number, it catches genuine changes in a ticker's risk profile that volatility
alone would overlook.
4. What is the information-theory score (mutual information)?
Information theory is the math behind how much knowing one thing tells you about another. The "mutual information" score measures shared structure between two stretches of data without assuming they move in a neat straight line. I compare a stock's most recent block of months with the block just before it and ask: how similar is the underlying pattern? A high score means the market's behavior is carrying forward smoothly; a sharp drop toward zero means the pattern just broke — the rhythm the stock had been following has changed. Unlike ordinary correlation, which tends to fail exactly when markets turn, this score keeps working during turbulence, making it a clean signal that a ticker may be entering a new phase.
5. What is a Hidden Markov Model (HMM)?
A Hidden Markov Model assumes the market is always in one of a few invisible
"moods." I use three: Bull (rising), Bear (falling), and Sideways
(drifting). I can't see the mood directly, but we can estimate the odds of each
from how returns have actually behaved — and one mood tends to lead into the next. For every ticker, the model reports the probability it's currently in each state,
for example "70% Bull, 20% Sideways, 10% Bear." Because markets usually stay in a mood for a while before flipping, knowing the current regime — and watching those percentages shift — helps us judge whether a stock's recent trend is likely to
continue or reverse.
6. What is the Google TimesFM AI model?
TimesFM is an artificial-intelligence model built by Google — the same kind of
"foundation model" technology behind today's AI chatbots, but trained to
forecast numbers over time instead of write sentences. It learned general
patterns by studying an enormous library of real-world data sequences, so it can
look at a stock's history of monthly returns and predict the most likely next
month. I feed it each ticker's track record and record its one-month-ahead
forecast. Its strength is spotting subtle, repeating patterns that fixed formulas
can't, giving us an independent, data-driven "AI opinion" that I weigh alongside
every other indicator.
7. What is the CBOE Skew Index?
The CBOE Skew Index measures how worried options traders are about a rare but
severe market crash — what people call "black swan" risk. It's built from the
prices investors pay for crash protection: when demand for that insurance rises,
Skew climbs. A normal reading sits around 100–115; unusually high readings mean the professionals are paying up to guard against a sudden plunge. I track how stretched Skew is versus its own recent history and treat spikes as a caution
flag — a sign the market is quietly pricing in elevated crash risk, which can
temper an otherwise bullish outlook.
8. What is the Commitments of Traders (COT) report?
Each week, the U.S. futures regulator publishes the Commitments of Traders report, showing how the biggest players are positioned in markets like the S&P 500, gold, and oil. I focus on two groups: speculators (hedge funds that chase
trends) and commercials (producers and hedgers who often lean against the
crowd). By checking whether speculators are unusually betting up or down compared with the past two years, I read the mood of the crowd. Extreme, one-sided bets often mark a trade that's become overcrowded and ripe for a reversal — useful for judging whether a trend still has fuel or is running on empty.
9. What is the Shiller CAPE ratio?
Created by Nobel laureate Robert Shiller, the CAPE ratio is a valuation
yardstick: it divides the market's price by its average inflation-adjusted
earnings over the past ten years, smoothing out short-term booms and busts. Think of it as a price tag telling you whether stocks are cheap or expensive compared with history. A high CAPE means investors are paying a lot for each dollar of earnings — historically a sign of weaker long-term returns ahead — while a low CAPE points to better value. I measure how far today's CAPE sits from its own long-run normal to gauge whether the overall market is stretched or attractively priced, which sets the backdrop for individual tickers.
10. What is RSI (14-month)?
The Relative Strength Index, or RSI, is a momentum gauge on a 0–100 scale that compares how strong a stock's recent gains have been against its losses. I run it over a 14-month window, so it reflects longer-term momentum rather than day-to-day noise. Readings above about 70 suggest a stock is "overbought" — it's climbed hard and may be due for a pause — while readings below about 30 suggest it's "oversold" and could be set for a bounce. It's a quick read on whether a ticker's trend is healthy or overstretched, and when enthusiasm (or fear) may have gone too far.
11. What is Beta?
Beta measures how much a stock tends to move relative to the overall market.
The market itself has a beta of 1.0: a stock with a beta of 1.5 typically swings
about 50% more than the market — both up and down — while a beta of 0.5 moves
about half as much. High-beta stocks (often tech and growth names) amplify rallies and sell-offs alike; low-beta stocks (often utilities and consumer staples) ride steadier. Knowing a ticker's beta tells you how aggressive or defensive it is, so
you can separate how much of its expected move is just the market's tide versus
something specific to that company.
Nothing here is investment advice — these indicators describe how I form my
views, not a recommendation to buy or sell any security.