Long-term investors must research and choose ETFs and stocks that compound on a straight line, for increasing log(Price) versus year. This means the slope, or return, must be the constant growth rate, which must be larger than the interest rate of bank deposits.
However, return is never a constant rate as treasury bills or bank deposits. You see many big bumps on log(Price) curve. The bumps up make you buy high. The bumps down make you sell low. That is how novices loose their money to the experts, or to the machines and algorithms.
Take a careful look at the 4 typical examples on the right: QQQ for Nasdaq-100, ARKK for active ETF on innovations, AAPL for mega cap stock, JNJ for value stock.
Gain is never constant, messy, but you can see a rough sine wave about a near horizontal line for the long-term growth rate. Can you see a period of 2 years? Switch back to log(Price) curve, you can now recognize the sine wave that swings about a linear axis with slope = growth rate. It is hard to see gain sine wave on log(Price) curve, because your eyes are pulled to the big bumps up/down. But the gain sine wave, with period = 2 years, approximates well the big slopes up/down over the years.
Does that mean we can extrapolate from this gain sine wave to find next year gain, or return, to decide what to invest in? Yes, you can predict return at 0, +6 months, or price at +6, +12 months, shown as red dots. No, the prediction error is larger the farther into the future you predict, especially in fast moving markets.
You must do further research on the fundamentals, to convince yourself that the growth rate and sine wave are not going to change a lot, because fundamentals are still good, even better? We will use these price and gain curves to convey our unbiased measurements and predictions, to help you invest more safely.