Interest rates are confusing, despite their ubiquity. This post takes an in-depth look at why interest rates behave as they do.

Understanding these concepts will help understand finance (mortgages & savings rates), along with the omnipresent e and natural logarithm. Hereβs our cheatsheet:

Term | Formula | Description & Usage |
---|---|---|

Simple | Fixed, non-growing return (bond coupons) | |

Compound (Annual) |
Changes each year (stock market, inflation) | |

Compound (n times per year) |
Changes each month/week/day (savings account) | |

Continuous Growth | Changes each instant (radioactive decay, temperature) | |

APR | Annual Percentage Rate (compounding not included) | |

APY | Annual Percentage Yield (all compounding effects included) |

- P = principal, your initial investment (i.e., $1,000)
- r = interest rate (i.e., 5% per year)
- n = number of time periods (i.e., 3 years)

And a quick calculator to convert APR to APY:

## Why the fuss?

Interest rates are complex. Like Roman numerals and hieroglyphics, our first system βworkedβ but wasnβt quite ideal.

In the beginning, you might have had 100 gold coins and were paid 12% per year (percent = per cent = per hundred β those Roman numerals still show up!). Itβs simple enough: we get 12 coins a year. But is it really 12?

If we break it down, it seems we earn 1 gold a month: 6 for January-June, and 6 for July-December. But wait a minute β after our June payout weβd have 106 gold in July, and yet earn only 6 during the rest of the year? Are you saying 100 and 106 earn the same amount in 6 months? By that logic, do 100 and 200 earn the same amount, too? Uh oh.

This issue didnβt seem to bother the ancient Egyptians, but did raise questions in the 1600s and led to Bernoulliβs discovery of e (sorry math fans, e wasnβt discovered via some hunch that a strange limit would have useful properties). Thereβs much to say about this riddle β just keep this in mind as we dissect interest rates:

**Interest rates and terminology were invented before the idea of compounding.**Heck, loans were around in 1500 BC, before exponents, 0, or even the decimal point! So itβs no wonder our discussions can get confusing.**Nature doesnβt wait for a human year before changing**. Interest earnings are a type of βgrowthβ, but natural phenomena like temperature and radioactive decay change constantly, every second and faster. This is one reason why physics equations model change with βeβ and not β(1+r)^{n}β: Nature rudely ignores our calendar when making adjustments.

## Learn the Lingo

As a result of these complications, we need a few terms to discuss interest rates:

**APR (annual percentage rate):**The rate someone tells you (β12% per year!β). Youβll see this as βrβ in the formula.**APY (annual percentage yield):**The rate you actually get after a year, after all compounding is taken into account. You can consider this βtotal returnβ in the formula. The APY is greater than or equal to the APR.

APR is what the bank tells you, the APY is what you pay (the price after taxes, shipping and handling, if you get my drift). And of course, banks advertise the rate that looks better.

Getting a credit card or car loan? Theyβll show the βlow APRβ youβre paying, to hide the higher APY. But opening a savings account? Well, of course theyβd tout the βhigh APYβ theyβre paying to look generous.

**The APY (actual yield) is what you care about, and the way to compare competing offers**.

## Simple Interest

Letβs start on the ground floor: **Simple interest pays a fixed amount over time**. A few examples:

- Aesopβs fable of the golden goose: every day it laid a single golden egg. It couldnβt lay faster, and the eggs didnβt grow into golden geese of their own.
- Corporate bonds: A bond with a face value of $1000 and 5% interest rate (coupon) pays you $50 per year until it expires. You canβt increase the face value, so $50/year what you will get from the bond. (In reality, the bond would pay $25 every 6 months).

Simple interest is the most **basic type of return**. Depositing $100 into an account with 50% simple (annual) interest looks like this:

You start with a principal (aka investment) of $100 and earn $50 each year. I imagine the blue principal βshovelingβ green money upwards every year.

However, this new, green money is stagnant β it canβt grow! With simple interest, the $50 just sits there. Only the original $100 can do βworkβ to generate money.

Simple interest has a simple formula: Every period you earn P * r (principal * interest rate). After n periods you have:

This formula works as long as βrβ and βnβ refer to the same time period. It could be years, months, or days β though in most cases, weβre considering annual interest. Thereβs no trickery because thereβs no compounding β interest canβt grow.

Simple interest is useful when:

**Your interest earnings create something that cannot grow more**. Itβs like the golden goose creating eggs, or a corporate bond paying money that cannot be reinvested.**You want simple, predictable, non-exponential results**. Suppose youβre encouraging your kids to save. You could explain that youβll put aside $1/month in βfun moneyβ for every $20 in their piggybank. Most kids would be thrilled and buy comic books each month. If your last name is Greenspan, your kid might ask to reinvest the dividend.

In practice, simple interest is fairly rare because most types of earnings can be reinvested. There really isnβt an APR vs APY distinction, since your earnings canβt change: you always earn the same amount per year.

## Really Understanding Growth

Most interest explanations stop there: hereβs the formula, now get on your merry way. Not here: letβs see whatβs really happening.

First, what does an interest rate mean? **I think of it as a type of βspeedβ**:

**50 mph**means youβll travel 50 miles in the course of an hour**r = 50% per year**means youβll earn 50% of your principal in the course of a year. If P = $100, youβll earn $50/year (your βspeed of money growthβ).

But both types of speed have a subtlety: **we donβt have to wait the full time period!**

Does driving 50 mph mean you must go a full hour? No way! You can drive βonlyβ 30 minutes and go 25 miles (50 mph * .5 hours). You could drive 15 minutes and go 12.5 miles (50 mph * .25 hours). You get the idea.

Interest rates are similar. An interest rate gives you a βtrajectoryβ or βpaceβ to follow. If you have $100 at a 50% simple interest rate, your pace is $50/year. But you donβt need to follow that pace for a full year! If you grew for 6 months, you should be entitled to $25. Take a look at this:

We start with $100, in blue. Each year that blue contributes $50 (in green) to our total amount. Of course, with simple interest our earnings are based on our original amount, not the βnew totalβ. Connecting the dots gives us a trendline: weβre following a path of $50/year. Our payouts look like a staircase because weβre only paid at the end of the year, but the trajectory still works.

**Simple interest keeps the same trajectory:** we earn βP*rβ each year, no matter what ($50/year in this case). That straight line perfectly predicts where weβll end up.

The idea of βfollowing a trajectoryβ may seem strange, but stick with it β it will really help when understanding the nature of e.

One point: the trajectory is βhow fastβ a bank account is growing at a certain moment. With simple interest, weβre stuck in a car going the same speed: $50/year, or 50 mph. In other cases, our rate may change, like a skydiver: they start off slow, but each second fall faster and faster. But at **any instant**, thereβs a single speed, a single trajectory.

(The math gurus will call this trajectory a βderivativeβ or βgradientβ. No need to hit a mosquito with the calculus sledgehammer just yet.)

## Basic Compound Interest

Simple interest should make you squirm. **Why canβt our interest earn money?** We should use the bond payouts ($50/year) to buy more bonds. Heck, we should use the golden eggs to fund research into cloning golden geese!

**Compound growth means your interest earns interest**. Einstein called it βone of the most powerful forces in natureβ, and itβs true. When you have a growing thing, which creates more growing things, which creates more growing thingsβ¦ your return adds up fast.

The most basic type is period-over-period return, which usually means βyear over yearβ. Reinvesting our interest annually looks like this:

We earn $50 from year 0 β 1, just like with simple interest. But in year 1-2, now that our total is $150, we can earn $75 this year (50% * 150) giving us $225. In year 2-3 we have $225, so we earn 50% of that, or $112.50.

In general, we have (1 + r) times more βstuffβ each year. After n years, this becomes:

Exponential growth outpaces simple, linear interest, which only had $250 in year 3 (100 + 3*50). Compound growth is useful when:

**Interest can be reinvested**, which is the case for most savings accounts.**You want to predict a future value based on a growth trend**. Most trends, like inflation, GDP growth, etc. are assumed to be βcompoundableβ. Yearly GDP growth of 3% over 10 years is really (1.03)^{10}= 1.344, or a 34.4% increase over that decade.

## Interest as a Factory

The typical interpretation sees money as a βblobβ that grows over time. This view works, but sometimes I like to see interest earnings as a βfactoryβ that generates more money:

Hereβs whatβs happening:

**Year 0:**We start with $100.**Year 1:**Our $100 creates a $50 βbondβ.**Year 2:**The $100 generates another $50 bond. The $50 generates a $25 bond. The total is 50 + 25 = 75, which matches up.**Year 3:**Things get a bit crazy. The $100 creates a third $50 bond. The two existing $50 bonds make $25 each. And the $25 makes a 12.50.**Years 4 to infinity:**Left as an exercise for the reader. (Donβt you love that textbook cop out?)

This is an interesting viewpoint. The $100 just mindlessly cranks out $50 βfactoriesβ, which start earning money independently (notice the 3 blue arrows from the blue principal to the green $50s). These $50 factories create $25 factories, and so on.

The pattern seems complex, but itβs simpler in a way as well. The $100 has no idea what those zany $50s are up to: as far as the $100 knows, weβre only making $50/year.

So whyβs this viewpoint useful?

**You can separate the impact of the parent ($100) from the children.**For example, at Year 3 we have $337.50 total. The parent has earned $150 (β3 * 50% * $100 = $150β, using the simple interest formula!). This means the various βchildrenβ have contributed $337.50 β $150 β $100 = $87.50, or about 1/3 the total value.**Breaking earnings into components helps understand e.**Knowing more about e is a good thing because it shows up everywhere.

And besides, seeing old ideas in a new light is always fun. For one of us, at least.

## Understanding the Trajectory

Oh, weβre not done yet. One more insight β take a look at our trajectory:

With simple interest, we kept the same pace forever ($50/year β pretty boring). With annually compounded interest, **we get a new trajectory each year**.

We deposit our money, go to sleep, and wake up at the end of the year:

**Year 1:**βHey, waittaminute. Iβve got $150 bucks! I should be making $75/year, not $50!β. You yell at your banker, crank up the dial to $75/year, and go to sleep again.**Year 2:**βHey! Iβve got $225, and should be making $112.50 per year!β. You scream at your bank and get the rate adjusted.

This process repeats forever β we seem to never learn.

## Compound Interest Revisited

Why are we waiting so long? Sure, waiting a year at a time is better than waiting βforeverβ (like simple interest), but I think we can do better. Letβs zoom in on a year:

Look at whatβs happening. The green line represents our starting pace ($50/year), and the solid area shows the cash in our account. After 6 months, weβve earned $25 but donβt see a dime! More importantly, after 6 months we have the same trajectory as when we started. The **interest gap** shows where weβve earned interest, but stay on our original trajectory (based on the original principal). Weβre losing out on what we should be making.

Imagine I took your money and returned it after 6 months. *βWell, ya see, I didnβt use it for a full year, so I donβt really owe you any interest. After all, interest is measured per year. Per yeeeeeaaaaar. Not per 6 months.β* Youβd smile and send Bubba to break my legs.

Annual payouts are man-made artifacts, used to keep things simple. But in reality, money should be earned all the time. We can pay interest after 6 months to reduce the gap:

Hereβs what happened:

- We start with $100 and a trajectory of $50/year, like normal
- After 6 months we get $25, giving us $125
- We head out using the new trajectory: 50% * $125 = $62.5/year
- After 6 months we collect 62.5/year times .5 year = 31.25. We have 125 + 31.25 = 156.25.

The key point is that our trajectory improved halfway through, and we earned 156.25, instead of the βexpectedβ 150. Also, early payout gave us a smaller gap area (in white), since our $25 of interest was doing work for the second half (it contributed the extra 6.25, or $25 * 50% * .5 years).

For 1 year, the impact of rate r compounded n times is:

In our case, we had (1 + 50%/2)^{2}. Repeating this for t years (multiplying t times) gives:

**Compound interest reduces the βdead spaceβ where our interest isnβt earning interest**. The more frequently we compound, the smaller the gap between earning interest and updating the trajectory.

## Continuous Growth

Clearly we want money to βcome onlineβ as fast as possible. Continuous growth is compound interest on steroids: you shrink the gap into oblivion, by dividing the year into more and more time periods:

The net effect is to make use of interest as soon as itβs created. We wait a millisecond, find our new sum, and go off in the new trajectory. Except itβs not every millisecond: itβs every nanosecond, picosecond, femtosecond, and intervals I donβt know the name for. Continuous growth keeps the trajectory perfectly in sync with your current amount.

Read the article on e for more details (e is a special number, like pi, and is roughly 2.718). If we have rate r and time t (in years), the result is:

If you have a 50% APR, it would be an APY of e^{.50} = 64.9% if compounded continuously. Thatβs a pretty big difference! Notice that e takes care of the icky parts, like dividing by an infinite number of periods.

Whyβs this useful?

**Most natural phenomena grow continuously**. As mentioned earlier, physical phenomena grows on its own schedule: radioactive material doesnβt wait for the Earth to go around the Sun before deciding to decay. Any physical equation that models change is going to use e^{rt}.**e**. It sounds strange, but e can even model the jumpy, staircase-like growth weβve seen with compound interest. Weβll get into this in a later article.^{rt}is the adjustable, one-size-fits-all exponential

Most interest discussions leave e out, as continuous interest is not often used in financial calculations. (Daily compounding, (1 + r/365)^{365}, is generous enough for your bank account, thank you very much. But seriously, daily compounding is a pretty good approximation of continuous growth.)

The exponential e is the bridge from our jumpy βdelayedβ growth to the smooth changes of the natural world.

## A Few Examples

Letβs try a few examples to make sure itβs sunk in. Remember: the APR is the rate they give you, the APY is what you actually earn (your true return).

**Is a 4.5 APY better than a 4.4 APR, compounded quarterly?**You need to compare APY to APY. 4.4% compounded quarterly is (1 + 4.4%/4)^{4}= 4.47% , so the 4.5% APY is still better.**Should I pay my mortgage at the end of the month, or the beginning?**The beginning, for sure. This way you knock out a chunk of debt early, preventing that βdebt factoryβ from earning interest for 30 days. Suppose your loan APY is 6% and your monthly payment is $2000. By paying at the start of the month, youβd save $2000 * 6% = $120/year, or $3600 throughout a 30-year mortgage. And a few grand is nothing to sneeze at.**Should I use several small payments, or one large payment?**. You want to pay debt off as early as possible. $500/week for 4 weeks is better than $2000 at the end of the month. Each payment stops a few weeksβ worth of interest. The math is a bit tricker, but think of it as 4 $500 investments, each getting different return. In a month, the first payment saves 3 weekβs worth of interest: 500 · (1 + daily rate)^{21}. The next saves 2 weeks: 500 · (1 + daily rate)^{14}. The third saves a week 500 · (1 + daily rate)^{7}and the last payment doesnβt save any interest. Regardless of the details,**prepayment will save you money.**

**The general principle:** When investing, get interest paid early, so it can compound. When borrowing, pay debt early to *prevent* that interest from compounding.

## Onward and Upward

This is a lot for one sitting, but I hope youβve seen the big picture:

**The interest rate (APR) is the βspeedβ at which money grows**.**Compounding lets you adjust your βspeedβ as you earn more interest**. The APR is the initial speed; the APY is the actual change during the year.**Man-made growth uses (1+r)**, or some variant. We like our loans to line up with years.^{n}**Nature uses e**. The universe doesnβt particularly care for our solar calendar.^{rt}**Interest rates are tricky.**When in doubt, ask for the APY and pay debt early.

Treating interest in this funky way (trajectories and factories) will help us understand some of eβs cooler properties, which come in handy for calculus. Also, try the Rule of 72 for a quick way to compute the effect of interest rates mentally (that investment with 6% APY will double in 12 years). Happy math.

## Leave a Reply

67 Comments on "A Visual Guide to Simple, Compound and Continuous Interest Rates"

Your articles are excellent, I just wait for the new ones to arrive and as I soon a new one appears, I am all excited to read.

Please keep going, thanks for all the better explanations.

Srikanth

Just a quick comment, I love your site and your articles… But here’s something I noticed…

When you’re stating “return = (1 + r/t)^tn” you need to make sure that you put the Principle in there :).

Return = P(1+r/t)^tn

This had me hammered for a little while as I tried to work out where we put the principle.

Admittedly I may just have been slow… π

@Srikanth: Appreciate the comment! Glad you’re finding the articles useful.

@Zachary: Thanks, that’s a good catch — just fixed it :). Nope, you’re right, it’s important to get those types of details correct.

[…] A Visual Guide to Simple, Compound and Continuous Interest Rates | BetterExplained (tags: money finance howto) […]

Dear Sir:

In the same breath, can you explain how scientists arrive at half life of radio-active material.

Regards,

T.Gopalan

another fun one! I love the way you simplify concepts without being condescending. Also, you leave ‘room’ for me to go off and find out more now that I have the basics down (basics with a different insight). The idea of ‘interest’ comes everywhere. For example, if you were in charge of cash flow at a factory. The ‘interest’ could be considered the ROI (return on investment) of the widget. If you had one dollar (it is a small factory) would you be better off paying the factory bills which amount to $1 or buying raw materials for the factory. Well let’s say you could buy materials for $1 and produce a gaggle of widgets which you can sell in the prescribed period of say 14 days. You can get payment in (I know lots of assumptions) in 25 days AND the bills are not due till net 30. Now the profits of that principal is $3. Cash flow being a function of interest aka time… did this have anything to do w/ your article???

later

T

[…] A Visual Guide to Simple, Compound and Continuous Interest Rates | BetterExplained (tags: finance money interest howto reference **mathematics) […]

Thanks Mr. Rose! Yes, that’s an interesting application — if you have the choice between paying bills later (in 30 days) or investing in your business today, the investment is a good idea because it will give you more breathing room down the road. And then you can reinvest those profits to become even more profitable.

@T.Gopalan: Hi, good question. Radioactive half-life is a bit different: rather than giving an interest rate, it’s more like an interest “time” — i.e., how long does it take for a material to decay to half it’s value?

This would be like giving interest rates in terms of the time needed to double: rather than 10% interest, you’d say “I have a doubling rate of 7.2 years” (or “halving rate” if your material is decaying by 10%/year).

I imagine expressing half life this way is useful because you often want to know the time, for figuring out things like carbon dating. In another article I’ll discuss how to convert from one to the other, but check out the Rule of 72 for more info.

Great article. APR versus APY is a distinction I’ve forgotten repeatedly, this should help it stick much better..

Thanks, glad you liked it! Yeah, I wish someone had told me earlier that APY is the rate you usually care about (“price after tax, shipping and handling”).

Kalid – again, I’m so impressed! I can’t tell you the number of times I’ve looked at interest rates and had no clue what it really meant. The way you structured the article was also very helpful to understanding the main points. I did notice the many e plugs, so I guess I’ll try to tackle a concept that has evaded me since high school. thanks!!

Nice article, Kalid. I think a more theoretical article would be useful too — one that discussed why interest rates exist at all and how they relate time and money.

Hi Jonah, thanks for dropping by. I agree, I think that’d be a great follow-up article, appreciate the suggestion.

@Desi: Awesome, so glad you liked it π

This is just brilliant. I have never quite understood e but this is so soooo simple. Even an old fella like me can learn something. This is a style of teaching that should be compulsory in all areas. Talk to me not at me and get results. Thanks heaps.

You’re welcome Stephen, I’m glad you enjoyed it. Rest assured, there’s plenty of young fellas (myself included) who went through the motions of e without really understanding what it was about :).

Hi Kalid,

This and the other articles on this site are really amazing!

It really explains concepts better than what they do in schools!

Take care and God bless!

Thanks, glad you’re enjoying the site! :).

Hi Kalid! This article is excellent!! I really enjoy your slogan philosophy: “Learn right, not rote.” Too often I think we’re shortchanged with understanding…

Just a quick catch of my eye:

(1) In your “Compound Interest” graph (the one that’s all blue), I believe the last column should be $337.50 not $327.50.

(2) If you do a quick find for the text “For a 1 year, the impact of rate r compounded t times is:”, I believe the “a” should be removed.

Your website is an inspiration and I look forward to learning more form you!

Thanks so much! π

Hi Krishna, thanks for the comment! I just fixed up the two typos, appreciate the pointers :).

I’m really happy you’re finding the site useful, it’s nice to interact with people who have a similar take on learning (that it can be about understanding and not just facts).

Again, thanks for the comment!

I’m a college finite math student and I was wondering if you could help me figure out how to convert a continuous rate into APY. In the way that something like 7.23% at a continuous rate is equal to APY. I’m trying to pick this up through my math book and can not figure it out at all.

Thanks

Hi Andrew, great question. When you have a 7.23% continuous rate, if you start with $1 at the end of the year you have

$1 * e^(.0723) = 1.07497778903 ~ 1.075

So 7.5% is your effective “APY”, that is your increase from the beginning of the year. Hope this helps!

Thank you for another great article that does as intended. I hope mathematicians will get the idea and follow suit.

There are two things I’d like to suggest, as they occurerd to me:

1) You show what happens if you compound twice in one year, how each additional compounded total itself adds to the next total to be compounded. The question I had, and I felt some students might have, is, what do you do if you want to make sure that at the end of the year the total interest does not exceed, say 50%, no matter how often it is compounded during the year. If at six months the interest of 50% makes the end result higher than a total of 50%, to make $131.25 (from the above examples), what percent would the halfway mark need to be to make the end result equal $150, i.e. 50% interest on P=$100?

2) What would the explanation and especially the equation be for a savings account that has an additional constant amount of money added per month, e.g. Start with $100, know the interest rate is 3%, but each month, besides the compounded interest, another $25 is added from each additional paycheck?

Thanks

@Armin: Glad you enjoyed it, those are great questions.

1) This is where the natural log comes into play. If you want a “final” return of 50% (that is, 1.00 becomes 1.50) you can do ln(1.50) = .405.

That means an interest rate of 40.5% will get you 50% return if it’s compounded as fast as possible. So, 40.5% is the “safe” rate you can use.

If you know you’ll only compound twice, then you can solve the equation:

(1 + r)^2 = 1.5

(1 + r) = sqrt(1.5) [take square root]

r = sqrt(1.5) – 1 [subtract 1]

r ~ .22

So, if you only compound twice (halfway & end of year) then you are safe with a rate of 22%.

2) That’s a really good question. The formula is a bit more complicated because you have to account for each deposit, which come in at different times. There’s more info here:

http://www.maths.leeds.ac.uk/Applied/0380/savings04.pdf

I think that would make a good follow-up article, as a similar formula is used to calculate loan payments (you pay the same amount each month for the loan, but how to do they work out that number given the interest?).

I have this question:

How does ‘e’ relate to compound interest?

Happy holidays!