5.4 Financial and Real-World Modeling

Modeling is a translation task

Financial and real-world questions are where Math 30-2 stops looking like isolated function practice. A diploma item may give a table, graph, short scenario, regression equation, or written-response prompt and expect you to choose a model, calculate a value, and interpret it. The official Mathematics 30-2 bulletin notes that students are generally successful with regressions but need to use non-rounded regression values for prediction, and that contextual problems are harder when the equation, table, or graph is not supplied. Those comments point to the same skill: translate before calculating.

The translation starts with units. If time is in years, use a yearly rate or convert the rate to match the compounding period. If time is in months, make sure the exponent counts months or that the model clearly divides by 12. If the output is dollars, people, hours of daylight, or degrees, the final sentence should use that unit. A number without context is not a complete modeling answer.

Choosing the model

Use the pattern in the context before choosing a formula. Constant amount added each step suggests a linear model. Constant percent change or repeated multiplication suggests an exponential model. A repeating rise-and-fall pattern suggests a sinusoidal model. A logarithmic model often appears as the inverse of exponential growth, especially when the question asks for the input or time needed to reach a target. Regression is appropriate when data points show a trend but no exact rule is provided.

Compound interest

For annual compounding, A = P(1 + r)^t, where P is principal, r is the annual interest rate as a decimal, and t is time in years. If interest is compounded m times per year, use A = P(1 + r/m)^(mt). The expression r/m is the interest rate per compounding period, and mt is the number of compounding periods.

Suppose $2,500 is invested at 4.2% per year compounded quarterly for 6 years. The quarterly rate is 0.042/4 = 0.0105, and the number of quarters is 4 * 6 = 24. The model is A = 2500(1.0105)^24, which gives about $3,212. The exponent 24 is not the number of years; it is the number of times interest is applied.

Thresholds and logarithms

If the same investment question asks when the account will first exceed $3,500, set 2500(1.0105)^n = 3500, where n is quarters. Then n = log(3500/2500) / log(1.0105), about 32.2 quarters. Since interest is applied quarterly, the account first exceeds the target after the 33rd quarter, or after 8.25 years. If the question asks for nearest tenth of a year instead, convert before rounding.

Regression and real data

A graphing calculator may produce an exponential, logarithmic, linear, quadratic, or sinusoidal regression equation. Do not copy rounded coefficients from the screen into every later step if the technology allows storing the regression equation. Predictions made with rounded coefficients can drift, especially when the input is far from the data. The bulletin's warning about non-rounded regression values is practical exam advice. In written response, state the model type, use enough precision, and explain whether the prediction is interpolation within the data or extrapolation beyond it.

Worked example: seasonal demand

A campground's electricity demand peaks at 1800 kWh in July and reaches a low of 900 kWh in January. The pattern repeats each year. A sinusoidal model is reasonable because the demand is seasonal. The amplitude is (1800 - 900) / 2 = 450 kWh, the midline is (1800 + 900) / 2 = 1350 kWh, and the period is 12 months. A model can be built from a July maximum, but a question may only ask for these features and their meanings. The amplitude means demand varies about 450 kWh above or below the average seasonal level.

Written-response habits

A strong response defines variables, writes the model, substitutes values, keeps unrounded calculator results, rounds only as directed, and concludes in context. It also notes limitations. An exponential investment model assumes the stated rate continues. A depreciation model may stop being realistic after the object reaches salvage value. A sinusoidal daylight model is useful for seasonal patterning but not for every local weather effect. These comments do not replace calculation, but they show mathematical understanding when the prompt asks to explain or justify.

The final trap is choosing a model from a keyword alone. "Growth" is not always exponential; a savings plan with a fixed monthly deposit may be linear if interest is ignored. "Seasonal" is not always sinusoidal if the data are irregular. Let the pattern and units decide.

Rounding and answer form

Financial answers often require a different rounding decision than pure function questions. Money is usually rounded to the nearest cent unless the prompt asks for dollars, while time-to-target questions may need the next whole compounding period because interest is applied only at the end of a period. State the rounded value and the reason for the rounding so the response matches the real situation.