Risk management

Volatility is a measure of how much a commodity’s price “swings” up and down. It describes the bumpiness of the ride.
High Volatility: Prices change rapidly and dramatically, like a rollercoaster. This creates uncertainty for budgets, margins, and cash flows.
Low Volatility: Market movements are subdued and prices stay relatively stable.

It calculates historical volatility by looking at past price data (e.g., the last 100 days) to see how “bumpy” that period was. In the software, you can measure annualized volatility for up to 20 symbols over a period you choose. This number then serves as a key input to the Value-at-Risk (VaR) simulation. The volatility calculator calculates start date to end date volatility.

: It follows a simple, two-step mathematical process:
Standard Deviation: The tool measures daily price changes (returns) over a set period, such as 90 days. It then calculates the standard deviation of those returns to measure how much a typical day’s return deviates from the average.
Annualization: To compare different assets, daily volatility is converted to an annual figure using the square root of time formula:

Annual \ Volatility = (Daily \ Standard \ Deviation) \times \sqrt{252}, weekly sqrt (52), monthly sqrt (12) and monthly average (sqrt 12 – for CFaR)
(Note: 252 represents the approximate number of trading days in a year).

Correlation measures how different commodity prices move in relation to each other. It is expressed as a coefficient between -1 and +1:
+1 (Positive Correlation): Commodities move in lock-step; when one goes up, the other tends to follow.
-1 (Negative Correlation): Commodities move in perfectly opposite directions, which is ideal for hedging.
0 (No Correlation): The price movements are random and unrelated.

It identifies how your selected commodities move together. This helps identify concentration risk (having too many assets moving the same way) and highlights diversification opportunities where combining assets might reduce overall risk.

VaR is a single number that quantifies potential downside. It answers: “What is the maximum expected loss for my exposure over a specific time, given a certain level of certainty?” Fastmarkets basis VaR off a notional value instead of mark-to-market, to allow the user flexibility in creating a VaR value or a Cashflow at Risk (CFaR) value.
Example: A 1-day, 99% VaR of $37,180 means there is only a 1% chance that your portfolio will lose more than $37,180 in the next day.

It is the “engine” of the VaR calculation.  The system runs thousands of simulated price paths based on your inputs to forecast potential risk exposure. It’s like rolling the dice thousands of times to see every possible market outcome instead of just guessing once.

The first step is Selecting Symbols. You can search and pick up to 20 Fastmarkets prices—such as different grades of aluminum or copper—to create a tailored view of your actual business exposure.

The software offers flexibility based on your expertise:
VaR Calculator – Auto: You pick symbols and weights; the system automatically computes volatility and correlations from history to estimate VaR.
VaR Calculator – Manual: You enter your own volatility and correlation figures to run custom “what-if” scenarios or stress tests.

Yes, significantly.
Short-Term (e.g., 1-day view): Shows immediate tactical risk.
Longer-Term (e.g., 20-day view): Shows how risk builds over longer periods, approximately one month.
The Risk: Using long-term “calm” data during a sudden geopolitical shock will cause the model to underestimate your current risk.

Correlation identifies the concentration of your prices. The correlation calculator tells you why your risk is so high, suggesting you may need to add offsetting commodities or derivatives to balance the portfolio and increase diversification.

Instead of being surprised by market swings, you quantify them. By understanding your “measurable number” of risk, you can stabilize production planning, protect margins, and make smarter hedging decisions than competitors who are only guessing.

VaR acts as a “Risk Alarm.” If the VaR simulation indicates a potential loss exceeding your risk appetite, it is a clear signal to hedge further. By looking at diversification results, you can see whether adding a specific, negatively correlated asset effectively brings the VaR back down to a comfortable level.

Any team exposed to raw-material volatility:
Procurement: To secure materials at predictable costs.
Treasury: To manage liquidity and forecast volatile cash flows.
Risk Management: To ensure that the company is functioning within risk limits

Yes. For a Treasurer, understanding price risk is directly tied to understanding cash flow volatility. By inputting your “notional value” (price multiplied by quantity), the VaR simulation estimates the potential hit to your actual cash reserves. You can change the weights from positive to negative to reflect revenues and costs.

The methodology in generating Notional VaR, Volatility and Correlation are as follows: 
Notional Value 
When you think of the starting point for measuring risk you need to know 
How much is at stake or how much money is my company exposed to? 
That number is your notional value (volume * price) 
By using notional value, a business can use the exact same mathematical logic to stress-test different scenarios: 
Your spending risk: The “Volumes” represent the quantity of raw materials you purchase, and the price is the cost of those materials 
Mark-to-Market (MtM): The “Volumes” represent current physical inventory or open spot positions. The price is the mark-to-market price 
Cash Flow at Risk (CFaR): The “Volumes” represent contracted future deliveries. The price is your costs or revenue prices. 
Earnings/Budget at Risk: The “Volumes” represent forecasted or budgeted production targets. 
Once the notional exposure is set and the historical price data is aligned the core risk metrics are applied. 
 
1. Volatility: The Measure of Movement 
Before assessing a portfolio, you must understand how wildly each individual price swings. Volatility is the baseline measure of historical uncertainty for a given price. 
The Concept: It is calculated using the historical standard deviation of the log returns (which represent the continuous period-over-period price returns) over a specific lookback period. It tells you how far, on average, the log returns of a price deviate from their average return. 
Time Scaling (Crucial Step): Volatility is typically quoted as an annualized percentage. Because VaR looks at shorter timeframes (e.g., 20 days), this annualized number must be scaled down using the Square Root of Time rule so the simulation doesn’t overstate the risk. 
2. Correlation: The Measure of Relationship 
Once you know how individual prices move, you must determine how they move together. Correlation is the mathematical engine behind portfolio diversification. 
The Concept: It compares the historical log returns of Price A and Price B day-by-day to see if they move in the same direction, opposite directions, or completely independently. The output is a number between -1.0 (moving in exact opposite directions—a perfect hedge) and 1.0 (moving in perfect lockstep). 
The Math (Pearson Correlation): The Pearson coefficient calculates this linear relationship by dividing the covariance of the two log returns by the product of their individual standard deviations. 
3. Value at Risk (VaR): The Measure of Potential Loss 
With the Notional Value, Volatility, and Correlation established, you can run a Monte Carlo simulation to forecast the maximum expected loss of the entire portfolio over a specific time horizon. 
The Process: 
Generate Scenarios: The system generates tens of thousands of random, normally distributed variables (simulating random future market shocks). 
Apply Relationships (Cholesky Decomposition): If you just applied random numbers, the prices would move independently. Running the random variables through a mathematical process called Cholesky Decomposition (the square root of the coefficient matrix) forces those shocks to respect your Correlation Matrix (e.g., if the simulation shocks one price, it knows exactly how to adjust a correlated price). 
Simulate Outcomes: Scale those correlated shocks by each price’s Time-Scaled Volatility, apply them to the Notional Value, and sum them up to create 20,000 (up to 1 million) simulated future portfolio values. 
Find the Worst Case: Sort all 20,000 outcomes from worst to best. To find your 95% Confidence VaR, you simply look at the 5th percentile outcome (the 1,000th worst scenario out of 20,000). 
This final number provides a single, digestible metric: under normal market conditions, there is a 95% probability the portfolio of price will not lose more than this specific currency amount over the given timeframe.