How Merkapital Quantifies Risk & Opportunity
The transparent logic that powers portfolios, probabilities, and decisions.
Investing is inherently uncertain — but your process shouldn't be. Merkapital's methodology quantifies probability, risk, and conviction using transparent, research-validated drivers. It's built to help you make defensible, repeatable decisions — not subjective guesses or opaque black-box outputs.
What This Methodology Enables
Instead of starting with math, start with outcomes:
- ✓Quantifies likelihood of positive returns over 1 year
- ✓Highlights asymmetric risk/reward profiles
- ✓Explains why each holding ranks where it does
- ✓Allows hypothesis testing with Scenario Lab
- ✓Supports defensible reporting for advisors
- ✓Produces portfolios that historically demonstrated risk-adjusted strength
Transparency & auditability
Every output is traceable: you can see factor scores, inputs from raw data, and how changes affect outcomes. This complements our disclaimers and supports advisor-level due diligence.
How It Works — Intuitive
Universe
~980 stocks
Scores
6 factors
Simulations
15k paths
Portfolio
Top 10
Stage 1 — Universe & Data Collection
We start with a broad, investable universe of ~980 stocks (S&P 500, large/mid-cap NYSE/NASDAQ, select ADRs). We collect high-quality fundamental inputs including:
- Earnings estimates
- Analyst price targets
- Profitability and ROE
- Valuation multiples
- Other factor data
We exclude micro-cap and OTC symbols to preserve data quality and reliability.
Why it matters: Good decisions start with good data. We never overwrite fresh data with stale inputs — ensuring reliability.
Stage 2 — Scoring & Simulation
Each stock is scored from 0–100 on six research-validated factors. These are not arbitrary weights — they reflect decades of peer-reviewed research on long-term expected returns.
For each stock, we estimate a distribution of outcomes by simulating thousands of potential price paths. This reveals not just a predicted price — but the range of plausible results and their likelihoods.
Outcome:
For each stock you get: Confidence Score (0–100), Probability of positive return (P(Up 1Y)), Median expected outcome, and realistic downside and upside ranges.
Why it matters: This is not a single forecast — it's a probability distribution. You see what can happen and how likely it is.
Portfolio Construction — Ranked, Not Random
Once every stock has a probability distribution, we rank all stocks by a risk-adjusted composite score, select the top 10, equal-weight each position, and validate the portfolio via 15,000 additional simulations.
Instead of complex optimization, we use equal weighting — a research-validated strategy that avoids unrealistic concentration and performs better under real-world uncertainty (DeMiguel, Garlappi & Uppal, 2009).
Why equal weight? It reduces estimation error and avoids overconfidence in any one forecast — a subtle but real source of risk in many quant portfolios.
View today's portfolio →Research-Backed Factors — Explained Simply
Each factor in plain language — and what it means for your decisions:
Earnings Growth
25%Measures forward profitability trends — one of the strongest predictors of returns.
Decision implication: High earnings growth increases the probability of upside capture.
Analyst Consensus
18%Gauges expectation gaps between price and consensus targets.
Decision implication: If the market underprices expected earnings upsides, expected outcomes improve.
Value (PEG)
18%Adjusts valuation for growth expectations.
Decision implication: Cheap growth has historically outperformed expensive growth.
Quality
18%High ROE and profit margins indicate sustainable profitability.
Decision implication: Quality companies exhibit better drawdown resilience.
Price Momentum
15%3-1 month return (skip last month). Winners tend to keep winning.
Decision implication: Momentum diversifies with value; combined signals improve risk-adjusted outcomes.
Market Base Rate
6%Anchors expectations to long-term historical returns.
Decision implication: Prevents overly bearish forecasts when markets are unstable.
Simulation Engine — What You Get
We simulate thousands of plausible return paths to estimate the full range of potential outcomes — not just a single forecast. To avoid overconfidence in forecasts, we temper extreme estimates using research-proven techniques — reducing the risk of unrealistic expectations.
How to Interpret the Output
- Confidence Score
- Strength of combined signals.
- P(Up 1Y)
- Likelihood of appreciation over one year.
- Median Return
- Central expectation.
- Downside / Upside Ranges
- Real risk/return boundaries.
These are not price targets — they are probabilities and ranges.
How It Appears in the Product
The methodology connects directly to what you see and use:
- •Confidence Score appears on the Stocks Dashboard and Portfolio pages.
- •Scenario Lab reuses the same scoring engine but allows custom assumptions (e.g., stress-test EPS or target prices).
- •Track Record uses the same simulations at the portfolio level — historical performance by confidence tier and vs SPY.
- •Reports include Monte Carlo output summaries, factor breakdowns, and rationale per holding.
News Context — Not Model Input Unless You Test It
We provide relevance-weighted news tone on each stock as context — not an automatic model driver.
You can test news-informed assumptions in Scenario Lab, which lets you stress test forward EPS, stress test target prices, and model alternative scenarios. Nothing changes unless you choose it.
Why this matters: Many tools use sentiment as a hidden signal. We treat it as hypothesis context, not model state.
Try Scenario Lab →Track Record & Proof
You have access to historical performance by confidence tier and portfolio outcomes vs SPY. This transparency ensures you can see whether these methodologies performed historically.
View Track Record →Institutional Report — Advisor-Ready Output
The output is not just a number. Our PDF export includes:
- • Factor breakdown per holding
- • Simulation output summary
- • Rationale paragraph per holding
- • Compare to benchmark
This is something an advisor can give to a client.
Generate Report →FAQs
Is this personalized advice?
No — this methodology interprets publicly available data to estimate probabilities. It does not account for personal financial circumstances. Merkapital is a probabilistic tool for informational purposes, not personalized investment advice.
Does the model guarantee future returns?
No — probabilities are not guarantees, but historically they correlate with observed outcomes. Past performance is not indicative of future results.
Does news sentiment automatically change forecasts?
No — only manual scenarios in Scenario Lab affect forecast assumptions. News tone is shown as context on each stock but does not feed into the model unless you run a scenario.
Why not optimization?
Estimation error makes equal weight more robust. Research (DeMiguel et al., 2009) shows that mean-variance optimization often underperforms simple equal weighting out-of-sample because return and covariance estimates are noisy. Equal weight reduces overconfidence in any single forecast.
Merkapital's methodology is built not for dogma, but for defensible decisions. By quantifying uncertainty, surfacing the drivers of outcomes, and enabling scenario testing, you can evaluate both what the model suggests and why it does so — and make investment decisions with transparent logic and measurable probabilities.