Tutorials

1Create Your Account

Head to the sign-up page and register with a username, email, and password (minimum 8 characters). Once registered, you are automatically logged in and your session persists across browser restarts.

2Open the Platform

Click Open Platform in the navigation bar or visit /app. You will see the main dashboard with the portfolio panel on top, risk metrics in the middle, and optimisation tools below.

3Add Your First Holding

In the portfolio panel, type a ticker symbol (e.g. AAPL) into the search field. Enter the number of shares and your average cost basis, then click Add. Soavalon fetches the latest market price and immediately displays your position value and unrealised P&L.

4Read Your Risk Metrics

Once you have at least two holdings, click Run Analysis. Soavalon computes:

• Portfolio Volatility — annualised standard deviation of your portfolio returns
• Value at Risk (VaR) — the maximum expected loss at a given confidence level
• Expected Shortfall — the average loss in the worst-case tail scenarios
• Sharpe Ratio — risk-adjusted return relative to the risk-free rate

5Save & Sync

Your portfolio is automatically saved to your browser's localStorage. When logged in, click Sync to Cloud to access your portfolio from any device.

1Portfolio Holdings Panel

The top section displays all your positions in a table with columns for ticker, shares, cost basis, current price, market value, and unrealised P&L. The summary row at the bottom shows total portfolio value and aggregate P&L. Click any column header to sort.

2Risk Configuration Panel

Below the holdings table you will find toggle buttons for configuring how risk is estimated:

• Volatility Mode — Sample (equal-weighted) or EWMA (λ = 0.94)
• Expected Return Mode — Historical CAGR, CAPM, Risk-Free, or Zero
• Lookback Window — 3 months to 5 years
• VaR Method — Parametric, Cornish-Fisher, or Historical

3Risk Results Panel

After clicking Run Analysis, results appear showing all three VaR methods side by side. The method you selected is highlighted, but all three are always visible for easy comparison. Below VaR you will see Expected Shortfall, portfolio volatility, and the correlation matrix.

4Optimisation & Stress Test

The bottom panels contain portfolio optimisation (Max Sharpe, Min Variance, Risk Parity, Equal Weight) and stress testing. These are covered in detail in the Intermediate and Advanced tutorials.

1Adding a Position

Type a valid ticker symbol into the search field (e.g. MSFT, TSLA, SPY). Enter the number of shares you hold and your average cost basis per share, then click Add. The position appears in your holdings table with live pricing.

2Editing & Deleting

Click the edit icon on any row to modify the share count or cost basis. Click the trash icon to remove the position entirely. All changes are auto-saved to localStorage.

3Understanding P&L

The P&L column shows your unrealised gain or loss for each position:

• P&L ($) = (Current Price − Cost Basis) × Shares
• P&L (%) = (Current Price / Cost Basis − 1) × 100

Green values indicate gains; red values indicate losses. The total P&L in the summary row aggregates all positions.

4Import & Export

Click Export to download your entire portfolio as a JSON file. To restore a backup or transfer between devices, click Import and select the JSON file.

5Named Portfolios

You can create multiple named portfolios and switch between them. Each portfolio maintains its own set of holdings, risk configuration, and optimisation results. Use this to compare different allocation strategies side by side.

1What Is Value at Risk?

VaR answers: “What is the maximum I can expect to lose over a given time horizon at a given confidence level?” For example, a 1-day 95% VaR of $10,000 means there is a 5% chance of losing more than $10,000 in a single day.

2Parametric VaR

Assumes portfolio returns are normally distributed. The formula is:

VaR = μ − z_α · σ

where z_α is the standard normal quantile (e.g. 1.645 for 95%, 2.326 for 99%). It is fast and widely used but underestimates tail risk for skewed or fat-tailed distributions.

3Cornish-Fisher VaR

Adjusts the normal quantile using the portfolio's skewness (S) and excess kurtosis (K):

z_CF = z + (z²−1)S/6 + (z³−3z)K/24 − (2z³−5z)S²/36

When the portfolio exhibits negative skew or positive excess kurtosis (fat tails), CF VaR will be larger than Parametric VaR — this is the correct behaviour, as it captures tail risk that the normal distribution misses.

4Historical VaR

Sorts the actual historical daily returns and picks the percentile corresponding to your confidence level. No distributional assumptions needed, but results depend entirely on the observed sample. If a particular type of market event did not occur in your lookback window, Historical VaR will not capture it.

5Choosing Confidence & Time Horizon

• 95% — standard for internal risk management
• 99% — used by Basel II/III for regulatory capital
• 99.5% — more conservative, captures extreme tails

Time horizons (1-day, 5-day, 10-day, 21-day) are scaled using the square-root-of-time rule: VaR_T = VaR₁ · √T.

1The Four Strategies

• Max Sharpe — finds the tangency portfolio with the highest excess return per unit of risk
• Min Variance — minimises portfolio volatility regardless of returns
• Risk Parity — equalises each asset's risk contribution (each provides 1/N of total variance)
• Equal Weight — 1/N allocation, a surprisingly strong baseline

2Setting Weight Constraints

Before running optimisation, set minimum and maximum weight bounds. For example, no position below 2% or above 40%. These constraints are enforced during the numerical solver, preventing concentrated bets.

3Cash Allocation

Optionally reserve a percentage in cash (earning the risk-free rate). The optimiser allocates the remaining capital across risky assets. This is useful for investors who want to maintain liquidity.

4Running the Optimisation

Select your strategy, configure constraints, and click Optimize. The results show:

• Optimal weights for each asset
• Expected portfolio return and volatility
• Sharpe ratio of the optimised portfolio
• A comparison with your current allocation

5Interpreting Results

Max Sharpe tends to concentrate in high-return assets, while Risk Parity produces more balanced allocations. Min Variance favours low-volatility assets. Compare all four to understand the trade-offs between return, risk, and diversification.

1Why Shrink Correlations?

Sample correlation matrices estimated from limited data can be noisy and poorly conditioned. This is especially problematic when the number of assets is large relative to the number of observations. Shrinkage blends the noisy sample estimate toward a structured target to produce a more stable matrix.

2The Shrinkage Formula

ρ_adj = w · ρ_sample + (1 − w) · ρ_target

• w = 1.0 — raw sample correlations (no shrinkage)
• w = 0.0 — full shrinkage to the target matrix
• Target — identity matrix (zero correlation) or constant-correlation matrix

Use the slider in the Risk Configuration panel to adjust the shrinkage weight.

3Beta Estimation Methods

• Daily Beta — OLS regression of asset returns against SPY returns over the lookback window
• Shrinkage (Blume-adjusted) — β_adj = 0.67β_raw + 0.33(1.0). Corrects for the empirical tendency of betas to mean-revert toward 1.0
• Market Beta — uses the beta reported by the data provider (Yahoo Finance) rather than computing from returns

4Practical Guidance

For portfolios with 5+ assets and lookback windows under 1 year, moderate shrinkage (w = 0.5 – 0.7) typically improves optimisation stability. Blume-adjusted beta is recommended for CAPM expected return estimates, as raw betas tend to overestimate risk for high-beta stocks and underestimate it for low-beta stocks.

1Navigating to Fundamentals

In the dashboard sidebar, click Fundamentals. The section has five sub-tabs: Calculators, Summary, Valuation Metrics, Earnings & Growth, and Balance Sheet. Enter any ticker at the top to load its financial data.

2CAPM Calculator

The Capital Asset Pricing Model estimates the expected return of an asset:

E(R) = R_f + β(E(R_m) − R_f)

Enter the risk-free rate, market return, and beta. You can also solve for any one variable by toggling the Solve For buttons. The calculator auto-fills beta when you enter a ticker.

3WACC Calculator

Weighted Average Cost of Capital blends the cost of equity and debt:

WACC = (E/V)R_e + (D/V)R_d(1 − T)

Input market cap (E), total debt (D), cost of equity (from CAPM), cost of debt, and tax rate. The result tells you the minimum return a company must earn to satisfy all capital providers.

4Relative Valuation & Risk Premium

• Relative Valuation — compares a company's P/E, P/B, and EV/EBITDA against sector peers to identify over- or under-valuation
• Risk Premium — estimates the equity risk premium (ERP) from historical market data or implied from current valuations

5Summary & Data Tabs

The Summary tab displays hero cards with current price, market cap, and beta, plus snapshot tables for valuation, earnings, and balance sheet highlights. The dedicated tabs go deeper:

• Valuation Metrics — trailing and forward P/E, P/B, P/S, EV/EBITDA, PEG ratio
• Earnings & Growth — EPS, revenue growth, profit margins, dividend yield and payout ratio
• Balance Sheet — total assets/liabilities, debt-to-equity, current ratio, cash flow, ROE, ROA, institutional ownership

1Setting Up a Comparison

Navigate to the Portfolio Comparison section in the Analysis tab. Select Portfolio A and Portfolio B from your saved portfolios using the dropdown selectors. Both portfolios must have holdings with historical price data available.

2Cumulative Return Chart

The comparison chart overlays the cumulative return paths of both portfolios over the selected lookback period. This gives an immediate visual picture of which portfolio has performed better and how their return profiles differ over time. Hover over the chart to see exact values at any date.

3Risk-Adjusted Metrics

Below the chart, a side-by-side metrics table compares:

• Sharpe Ratio — excess return per unit of total risk (higher is better)
• Sortino Ratio — excess return per unit of downside risk (penalises only negative volatility)
• Annualised Volatility — standard deviation of returns scaled to a year
• Max Drawdown — largest peak-to-trough decline during the period

4Probabilistic Dominance

Soavalon computes the probability that one portfolio outperforms the other over random sub-periods. If Portfolio A dominates with 70% probability, it means that in 70% of sampled time windows, A delivered higher risk-adjusted returns than B.

5Optimal Blend

The optimal blend feature finds the allocation split between A and B that maximises the Sharpe ratio. For example, the result might show “65% Portfolio A + 35% Portfolio B” as the optimal mix. Use this to understand whether combining strategies can improve your overall risk-return profile.

6Holdings Comparison

A detailed table lists every holding from both portfolios, showing which assets are unique to each and which overlap. Overlapping positions display the weight difference, helping you identify where the two portfolios diverge most.

1Understanding Factor Models

A factor model explains asset returns as a function of common risk factors. Soavalon's default uses two factors: SPY (broad market) and QQQ (tech/growth). Each asset's sensitivity to these factors is estimated via multivariate regression on historical returns.

2Two-Factor vs Single-Factor

• Two-Factor (SPY + QQQ) — captures both market-wide and tech-sector risk. Ideal for portfolios with significant tech exposure.
• Single-Factor (SPY only) — simpler model where each asset's beta to SPY determines its stress response. Good for broad, diversified portfolios.

3Configuring a Stress Scenario

Enter hypothetical percentage changes for each factor. For example:

• SPY: −10% (broad market crash)
• QQQ: −15% (tech selloff, worse than market)

Soavalon then estimates per-position P&L as: Position Value × (Factor Loading × Shock Magnitude).

4Reading the Results Table

The stress test output shows a breakdown for each asset:

• Factor Loadings — how sensitive the asset is to each factor
• Estimated $ Impact — dollar loss or gain per position
• % Impact — percentage change in each position's value

The total row aggregates all per-position impacts.

5Identifying Concentration Risk

If a single position dominates the total stress loss, your portfolio has concentration risk. Consider rebalancing or hedging that exposure. Compare results across different shock scenarios (mild, moderate, severe) to understand how your portfolio responds to varying levels of market distress.

1What Is Monte Carlo Simulation?

Monte Carlo simulation generates thousands of possible future return paths by randomly sampling from a statistical model. By aggregating outcomes across all paths, you can estimate the full distribution of portfolio returns — including rare tail events that parametric methods may miss.

2Geometric Brownian Motion (GBM)

Each asset's price path follows:

S_t+1 = S_t · exp((μ − σ²/2)Δt + σ√Δt · Z)

where Z is a standard normal random variable, μ is the expected return, and σ is the volatility. This ensures prices stay positive and returns are log-normally distributed.

3Correlated Multi-Asset Simulation

For a portfolio, asset returns must be correlated. Soavalon uses a Cholesky decomposition of the correlation matrix to generate correlated random variables:

• Compute the Cholesky factor L from the correlation matrix: Σ = LL^T
• Generate independent standard normal vectors Z
• Transform: correlated returns = L · Z

This preserves the pairwise correlation structure across all simulated paths.

4Extracting Tail Risk Percentiles

After running N simulations (e.g. 10,000), sort the terminal portfolio values. The 5th percentile gives the 95% VaR; the 1st percentile gives the 99% VaR. The average of all values below the VaR threshold gives the Expected Shortfall.

5Practical Considerations

• More simulations = more stable estimates (10,000 is a good starting point)
• GBM assumes constant μ and σ; real markets exhibit stochastic volatility and jumps
• Compare MC-derived VaR with Parametric and Historical VaR to cross-validate your risk estimates

1Opening the Options Pricer

In the dashboard sidebar, click Options. The section has two sub-tabs: Black-Scholes and Taylor Expansion. You can auto-fill the current stock price by entering a ticker at the top.

2Black-Scholes Inputs

Enter the five standard parameters:

• S — current stock price
• K — strike price
• T — time to expiration (in years)
• r — risk-free interest rate
• σ — volatility (annualised)

Select Call or Put, then choose what to solve for. The default solves for the option price, but you can toggle to solve for Implied Volatility (given a market price) or Stock Price (given a target option value).

3Interpreting the Greeks

After computing the price, Soavalon displays all five Greeks:

• Delta (Δ) — rate of change in option price per $1 move in the underlying. Calls: 0 to 1, Puts: −1 to 0
• Gamma (Γ) — rate of change of delta per $1 move. Highest for at-the-money options near expiry
• Theta (Θ) — time decay per day. Always negative for long options — your position loses this much value each day, all else equal
• Vega (ν) — sensitivity to a 1% change in implied volatility. Higher for longer-dated options
• Rho (ρ) — sensitivity to a 1% change in the risk-free rate. Usually a minor factor for short-dated options

4Quick Reference Sidebar

The right-hand panel shows a quick reference with the Black-Scholes formula, key relationships, and boundary conditions. Use this as a cheat sheet while experimenting with different inputs.

5Taylor Expansion for P&L Estimation

Switch to the Taylor Expansion tab to estimate how an option position’s value changes under scenario shifts. Enter your current position details (option price, quantity, Greeks), then specify hypothetical changes:

• dS — change in stock price
• dσ — change in implied volatility
• dt — time elapsed (in days)

The estimated P&L is computed using a second-order Taylor approximation:

dV ≈ Δ·dS + ½Γ·dS² + Θ·dt + ν·dσ

A contribution breakdown bar shows how much of the total P&L comes from each Greek, making it easy to identify which risk factor dominates.

6Practical Use Cases

• Hedging — use delta to determine shares needed to delta-hedge an option position
• Earnings trades — check vega to understand your exposure to post-earnings IV crush
• Decay management — monitor theta to decide when to roll or close short-dated positions