Something about the challenging nature of calculating compound interest keeps drawing me back, as in my *Mental Financial Wizard* post and my recent *Estimating Compound Interest* post. Or, maybe I'm just greedy.

In either case, here's yet another way to get a good estimate of interest compounded over time. It's a little tricky to do in your head alone, so you'll probably prefer to work this one out on a sheet of paper.

It turns out that compound interest is based on the binomial theorem. This means we can use relatively simple math concepts from Pascal's Triangle (also based on the binomial theorem). The method I'm about to teach you has its roots in the approach used to work out coefficients in *The Easy Peasy Binomial Expansion Trick* (jump down to the paragraph which reads, *"So now comes the part where the coefficients for each term are written. This is very easy to do with the way we set up our example."*).

What we're going to be estimating is the total percentage of interest alone. Once this is done, you can calculate the original investment into the problem. As a first example, let's work out 5% interest per year for 10 years. To keep things simple, we'll work with 5% as if it represented 5, instead of 0.05.

To get a starting point multiply the interest rate by the time, as if you were working out simple interest. In our 5% for 10 years example, we would simply multiply 5 × 10 = 50. You need to make a table with that number expressed two ways: As a standard number, as as a fraction over 1. For this example, the first row of the table would look like this:

Number | Fraction |
---|---|

50 | ^{50}⁄_{1} |

From here, there are 2 repeating steps, which repeat only as many times as you wish to carry them.

**STEP 1:**You're going to create a new fraction in the next row, based on the existing fraction. Take the existing fraction,

*increase*the denominator (the bottom number of the fraction) by 1, and

*decrease*the numerator by the amount of the annual interest.

In our example, starting from

^{50}⁄

_{1}, we'd increase the denominator by 1, turning it into

^{50}⁄

_{2}, and then decrease the denominator by 5, because we're dealing with 5% interest, to give us

^{45}⁄

_{2}. The table, in this example, would now look like this:

Number | Fraction |
---|---|

50 | ^{50}⁄_{1} |

^{45}⁄_{2} |

**STEP 2:**Take the number from the previous row, and multiply this by the new fraction, in order to get a new number for the current row. Divide the result by 100, and write this number down in the new row. This can seem challenging without a calculator, but if you think of a fraction as simply telling you to divide by the denominator and multiply by the numerator, it becomes simpler.

Continuing with our example, we'll multiply the number from the previous row (50) times our new fraction (

^{45}⁄

_{2}). That's 50 × 45 ÷ 2 = 25 × 45 = 1,125. 1,125 ÷ 100 = 11.25, so we add that number to the new row like this:

Number | Fraction |
---|---|

50 | ^{50}⁄_{1} |

11.25 | ^{45}⁄_{2} |

From here, we can repeat steps 1 and 2 as many times as we like, depending on what kind of accuracy is needed. Repeating step 1 one more time, we get this result (do you understand how we got to

^{40}⁄

_{3}?):

Number | Fraction |
---|---|

50 | ^{50}⁄_{1} |

11.25 | ^{45}⁄_{2} |

^{40}⁄_{3} |

After repeating step 2, we work out 11.25 × 40 ÷ 3 = 11.25 × 4 × 10 ÷ 3 = 45 × 10 ÷ 3 = 450 ÷ 3 = 150. Don't forget, as always, to divide by 100, which gives us 1.5 for the new row:

Number | Fraction |
---|---|

50 | ^{50}⁄_{1} |

11.25 | ^{45}⁄_{2} |

1.5 | ^{40}⁄_{3} |

Most of the time, I stop the calculations when the number in the bottommost row is somewhere between 0 and 10. I find this is enough accuracy for a decent estimate.

Once you've stopped generating numbers, all you need to do to estimate the interest percentage is add up everything in the the

*Number*column! In our above example, we'd add 50 + 11.25 + 1.5 to get 62.75. In other words, 5% for 10 years would yield roughly 62.75% interest. If we run the actual numbers through Wolfram|Alpha, we see that the actual result is about 62.89% interest. That's not bad for a paper estimate!

Back in 2012, a question was posted at math stackexchange which could've benefitted from this technique. In under 3 minutes, answer the following multiple choice question without using a calculator or log tables:

Someone invested $2,000 in a fund with an interest rate of 1% a month for 24 months. Consider it to be compounded interest. What will be the accumulated value of the investment after 24 months?Let's use this technique to work this out:

A) $2,437.53

B) $2,465.86

C) $2,539.47

D) $2,546.68

E) $2,697.40

Number | Fraction |
---|---|

24 | ^{24}⁄_{1} |

2.76 | ^{23}⁄_{2} |

0.2024 | ^{22}⁄_{3} |

Hmmm...24 + 2.76 + 0.2024 = 26.9624, so that would give us about a 26.96% return, or a little less than 27%. Multiplying this by 2,000 is easy, since we can multiply by 2, then 1,000. This lets us know there must be just under $540 in interest on that $2,000. A and B are way too low, E is way too high, and D is just over $540 in interest. That eliminates every answer except C. Sure enough, Wolfram|Alpha confirms that $2,539.47 is the correct answer!

## 1 Response to Estimating Compound Interest Without a Calculator

It�s difficult to find well-informed people about this subject, but you sound like you know what you�re talking about!

Thanks

Post a Comment