Help center

What a "factor" is

A factor is a measurable property of a stock — quality, momentum, volatility, value — that has historically been associated with a return premium relative to the broad market. Decades of academic research, replicated on Indian markets, have identified a stable set of factors that show up across cycles.

The six in the Markov stack

Quality

A stock is "high quality" if the underlying business shows consistent return-on-equity, manageable leverage and stable margins across years. The quality factor captures the long-running observation that high-quality businesses tend to outperform low-quality ones in risk-adjusted terms. Markov's persistent quality and QMJ strategies lean on this.

Momentum

Stocks that have outperformed recently tend to continue outperforming for some period before mean-reverting. This is the most widely replicated anomaly in finance. Sector-neutral momentum and QMJ both draw on it.

Residual momentum

A refinement on plain momentum. Strip out the broad-market and sector contributions from each stock's recent return; what remains is the stock-specific component. Sorting on residual momentum has historically given a cleaner signal than sorting on total momentum because it isolates the stock's story from the market's and the sector's.

Low volatility

Counter-intuitively, the lowest-volatility cohort of the stock market has historically delivered better risk-adjusted returns than the highest-volatility cohort. Low-vol quality combines this with the quality factor to bias toward smooth-riding businesses.

Composite (multi-factor)

QMJ — Quality / Momentum / Junk — combines a quality rank, a momentum rank and the inverse of a "junk" rank into a single composite score. Composites tend to be more robust than single-factor sorts because they're less likely to whipsaw on any one factor's bad year.

What "factor" doesn't mean

It doesn't mean a tip, a forecast or a guess. The factor concept is deliberately humble: it doesn't predict any specific stock's next move. It says: across a large basket, stocks scoring high on these properties have historically outperformed stocks scoring low on them.

Why a basket, not a single name

Factor strategies are cross-sectional. The edge shows up when you hold the top decile against the bottom decile (or just the top decile against the index) across hundreds of names. Picking one stock because it scores well on a factor is much weaker — the law of large numbers is what makes the factor work.

The Markov feed reflects this: a strategy publishes a list of names each rebalance, not one hero pick.

Reading the academic literature

Some accessible starting points if you want to dig in:

• Fama & French — the original three-factor model.
• Asness, Frazzini & Pedersen — "Quality Minus Junk" (the conceptual ancestor of our QMJ).
• Blitz & Vidojevic — low-volatility anomaly papers.

None of these are required reading to use the app, but they're a clean on-ramp if you want to understand why our methodology is the way it is.