Chapter 17 · Investment

HOST 00:00

Welcome back. Today we are on investment: in machines, in shares and in houses. In October 2026, American stocks cost about 40 times their average yearly profits, or earnings. That's close to the record set in 1999, at the peak of the dot-com boom.

ECONOMIST 00:19

The dot-com boom was a rush into internet shares in the late 1990s, before they crashed. So is this a bubble again? A bubble is a price far above what the future payments of a share or a house are worth today. Or are buyers right to pay so much? One simple tool helps answer that.

HOST 00:40

And that same tool also tells a boat builder whether to buy one more saw?

ECONOMIST 00:44

Yes. The tool is one comparison: money in the bank, or money in the thing you buy. That comparison prices a machine, a share and a house.

HOST 00:54

So here is today's question. When is a machine, a share or a house worth buying, and why does investment swing so much?

ECONOMIST 01:03

By the end of this episode, you'll be able to do three things. First, describe how investment falls in recessions, and explain why it swings so much. Second, find the yearly cost of using a machine, and show why a 35 percent tax on profits cuts the machines a firm wants by almost half. Third, explain how the same idea prices a share and a house, and why small changes in expected growth move those prices a lot.

HOST 01:32

And we'll follow the chapter's mind map part by part?

ECONOMIST 01:35

Yes. Part one looks at how investment swings in the data. Part two builds the user cost of capital, the yearly cost of using a machine. Part three uses the same idea for stocks, houses and policy. Each part is a branch of the map, with a few leaves.

HOST 01:53

Part one of the map: Investment swings. First, what do economists mean by investment? To me, investing means buying shares.

ECONOMIST 02:03

That's the everyday meaning, but not ours. Buying a share only moves ownership from one saver to another. Here, investment means spending on new capital, new things that help us produce more in the future.

HOST 02:17

And the first leaf says it has three parts.

ECONOMIST 02:20

Right. Business investment is the equipment, buildings, software and research of firms. Housing is the building of new homes. And inventories are the change in the goods that firms keep in store, made but not yet sold.

HOST 02:36

Next leaf: investment in 2006 and 2009. How big was the fall?

ECONOMIST 02:41

We measure investment as a share of GDP, a country's total output in a year. In the United States, private investment was 19.6 percent of GDP in 2006. By 2009 it was down to 13.3 percent.

HOST 02:56

So it fell by about 6 points of GDP. Let me guess which part fell most. Firms stop building factories in a recession, so my guess is business investment.

ECONOMIST 03:07

That's a natural guess, and it was right in the dot-com bust of 2000 to 2002. But this time it's wrong. These are percentage points, the plain difference between two percentages. Of the fall of about 6 points, housing alone explains about 3.4, more than half. Inventories and business investment share the rest.

HOST 03:30

Then the next leaf, housing, is the big story.

ECONOMIST 03:33

It is. Housing investment peaked at 6.6 percent of GDP in 2005. By 2011 it was down to 2.4 percent, less than half. That was the housing bust, a sharp fall after a boom, in the Great Recession. Part three shows why expected price gains matter so much for house prices.

HOST 03:55

Next leaf: inventories. In a recession, shops can't sell their goods. So unsold goods build up in the warehouses, and inventories rise.

ECONOMIST 04:05

A lot of students think that, but the data say the opposite. Inventory investment turned negative in recession years such as 1982, 2009 and 2020. Negative means firms took more goods out of store than they put in, so their inventories shrank.

HOST 04:22

Why would firms do that?

ECONOMIST 04:24

Because they cut production by more than their sales fall. Many inventories are parts and materials inside the production line. Less production means fewer parts in the line. So inventory investment rises in booms and falls in recessions.

HOST 04:39

Last leaf of this part: Ireland's housing. Was the housing boom only an American story?

ECONOMIST 04:44

No, and some booms were much bigger. In Ireland, housing investment reached 13.5 percent of GDP in 2006, about twice the U.S. peak. By 2011 it had fallen to 2 percent.

HOST 04:59

So in part one we saw that investment is the part of spending that swings most. In the Great Recession, housing and inventories fell hardest, and Ireland's housing fell even more. But why does investment swing so much?

ECONOMIST 05:14

Because every investment is a bet on the future. Next branch: how a firm decides how much capital to own.

HOST 05:23

Part two of the map: The user cost of capital. Where do we start?

ECONOMIST 05:29

On Islandia, our imaginary island country. Harbor Boatworks builds the island's fishing boats in a workshop. It has some spare cash. It can leave the money in a savings account at the bank in town, or buy one more power saw.

HOST 05:45

And the first leaf, the arbitrage equation, compares those two choices?

ECONOMIST 05:51

Exactly. Put the same dollars into each choice for a year. The bank pays interest. The saw pays the extra output it makes, plus any change in its price when the firm sells it. Economists call that extra output the marginal product of capital. When the firm has chosen well, both choices pay the same.

HOST 06:13

Why must they pay the same?

ECONOMIST 06:15

Suppose the saw paid more than the bank. Then the firm buys another saw. But each extra saw has less work to do, so its extra output is smaller. The firm keeps buying until the last saw pays just what the bank pays. Moving money toward the better choice until no gain is left is called arbitrage.

HOST 06:38

Next leaf: the user cost. What does a saw really cost the firm each year?

ECONOMIST 06:43

Three things. First, the interest the firm gives up by not leaving the money in the bank. Second, wear and tear, the part of the saw that is used up during the year. Third, take away any rise in the saw's price, because the firm gets that back when it sells. Together, that's the user cost of capital.

HOST 07:02

Give me Islandia's numbers.

ECONOMIST 07:04

The bank pays 2 percent a year, after inflation. Each year, 8 percent of a saw wears out. Saw prices stay the same. Add them up, and the user cost is 10 percent a year. Each dollar of saw must add at least 10 cents of output a year.

HOST 07:19

Now the tax leaf. The Islandia Treasury taxes Harbor Boatworks' profits. How does that change the user cost?

ECONOMIST 07:29

The Treasury takes a slice of what every saw produces. But it pays none of the saw's costs. So the saw must produce more before tax, to leave the firm the same return after tax.

HOST 07:43

How much more?

ECONOMIST 07:44

Let's build that number step by step. Say the tax rate is 35 percent. Then the firm keeps only 65 cents of every dollar a saw produces. After tax, it still needs 10 cents per dollar of saw.

HOST 07:58

So before tax, the saw must produce 10 cents divided by 0.65?

ECONOMIST 08:04

Yes, about 15.4 cents. So the user cost rises from 10 to 15.4 percent. In cents, 15.4 is about 54 percent more than 10.

HOST 08:16

Next leaf: desired capital, the number of saws the firm wants. A 35 percent tax means the firm wants 35 percent fewer saws.

ECONOMIST 08:25

That's the tempting answer, but the fall is bigger, almost half. To see why, ask how fast a saw's extra output rises when the firm owns fewer saws. In this model, it rises slowly. Suppose Harbor Boatworks had only half as many saws. The last saw's extra output would rise from 10 cents to only about 16 cents. That's a rise of about 59 percent, not double.

HOST 08:51

And the tax asks for a rise of about 54 percent, a bit less than 59. So the firm must cut its saws by a bit less than half?

ECONOMIST 09:01

Exactly. Here is the exact rule. The tax divides the user cost by 0.65, the share the firm keeps. Then the number of saws the firm wants is multiplied by 0.65 times the square root of 0.65.

HOST 09:16

And the square root of 0.65 is?

ECONOMIST 09:19

About 0.81. Now multiply: 0.65 times 0.81 is about 0.52. So the firm wants 0.52 times its old number of saws, about half. That's a fall of about 48 percent.

HOST 09:34

Does it sell half its saws all at once?

ECONOMIST 09:37

No. Investment is the flow that takes the firm from the saws it has to the saws it wants. With too many saws, it simply stops buying, and wear shrinks its capital.

HOST 09:48

So investment can swing much more than capital?

ECONOMIST 09:52

Yes, and this is why investment is so unstable. A firm that already has the saws it wants buys about 8 cents of new saws a year for each dollar of saws, just to replace wear. Now suppose it wants just 8 percent fewer saws. Then it can stop buying for a whole year. A small change in the capital it wants cuts this year's investment to zero.

HOST 10:16

Last leaf of this part: the investment rate. That's the share of output a country spends on new capital each year, right?

ECONOMIST 10:22

Yes. Take all of Islandia's firms together. Say their capital grows 2 percent a year. Add the 8 percent that wears out. So for each dollar of capital, firms need 10 cents of new capital a year.

HOST 10:36

And how do we turn that into a share of output?

ECONOMIST 10:38

As in the production model of Chapter 4, owners of capital get one third of output. Each dollar of capital earns its user cost, 10 cents. So output is three times that: 30 cents for each dollar of capital. The investment rate is the 10 cents of new capital divided by the 30 cents of output. That's one third, or 33.3 percent. Our round numbers make it higher than in the U.S. data.

HOST 11:04

And with the 35 percent tax?

ECONOMIST 11:06

Firms still need 10 cents of new capital for each dollar of capital. But now each dollar earns 15.4 cents, so it goes with about 46 cents of output. 10 divided by 46 is about 21.7 percent, down from 33.3. The Solow model of Chapter 5 simply assumed the investment rate. Here, taxes and interest rates explain it.

HOST 11:32

So in part two we saw that a machine must earn the bank's interest plus its wear, minus any price gain. A tax makes that target higher, and the firm wants far fewer machines.

ECONOMIST 11:45

And each year's investment is small compared with the capital firms own. So when rates, taxes or expected prices change the capital firms want a little, investment changes a lot. Next branch: stocks and houses.

HOST 12:01

Part three of the map: Stocks, houses and policy. First leaf: a stock's price. Islanders can buy shares of Harbor Boatworks. What is a share worth?

ECONOMIST 12:14

Same comparison: savings in the bank, or a share held for a year. The share pays a dividend, a part of the firm's profits paid to each owner. Then it sells at a new price. With no risk, both choices must pay the same.

HOST 12:28

Numbers, please.

ECONOMIST 12:29

For shares, we use new round numbers: the bank pays 5 percent. Say the share pays 3 dollars a year, forever, with no growth. Then the dividend alone must match the bank's 5 percent. So the price is 3 dollars divided by 5 percent: 60 dollars.

HOST 12:45

And if the dividend grows?

ECONOMIST 12:48

Let it grow 2 percent a year. The price then grows just as fast, a yearly price gain for the owner. So the dividend only needs to pay the other 3 percent: 3 dollars divided by 3 percent is 100 dollars. In general, the price is the dividend divided by the gap between interest and growth. Since that gap is small, small changes in it move the price a lot.

HOST 13:13

Next leaf: the price-earnings ratio. That's the share price divided by the firm's earnings, its profit per share?

ECONOMIST 13:21

Yes. The economist Robert Shiller divides the price by average earnings over the past ten years. For the U.S. stock market, his ratio was 40.7 in October 2026. Its average since 1881 is 17.8.

HOST 13:36

And what does our model say the ratio should be?

ECONOMIST 13:39

Divide the share price by earnings, and you get a simple rule. The ratio is the share of earnings paid out as dividends, divided by the gap. Say firms pay out half, and the gap is 3 percent as before. One half divided by 0.03 is about 16.7, close to the long-run average.

HOST 13:59

So today's 40.7 is more than twice the average. My guess: prices are twice too high. It's a bubble.

ECONOMIST 14:08

Maybe, but the model says we can't be sure. Turn the rule around. A ratio of 40 means the gap is one half divided by 40, only one and a quarter percentage points. With the bank at 5 percent, investors must expect dividends to grow about 3.75 percent a year, forever.

HOST 14:28

That sounds very optimistic.

ECONOMIST 14:30

It may be. But a small change in beliefs about growth or interest rates can justify a big change in prices. So a bubble is hard to tell apart from real optimism. The same debate ran in 1999, when the ratio hit its record, 44.2.

HOST 14:49

Next leaf: Tobin's q. What is q?

ECONOMIST 14:54

The economist James Tobin's q compares two values of a firm. The first is what the stock market says the whole firm is worth. The second is what it would cost to buy all its capital new. Tobin's q is the first divided by the second.

HOST 15:09

And when q is above 1, the firm should invest?

ECONOMIST 15:13

Yes. Then Harbor Boatworks could sell new shares, buy saws with the money, and be worth more than the saws cost. The market values a firm by what its machines will earn. So a q above 1 means a saw earns more than its user cost. Below 1, the firm is worth less than its machines, and it should shrink.

HOST 15:37

Next leaf: house price. Islanders buy seaside cottages with a mortgage, a loan to buy a house, and rent them out. Same idea?

ECONOMIST 15:46

Same idea. A house is worth its rent divided by its user cost. The buyer pays part of the price with her own money, the down payment, and borrows the rest.

HOST 15:58

So what is the user cost of a house?

ECONOMIST 16:01

First, the interest on the loan. Islandia's tax rules let owners subtract that interest from their taxable income, so it costs a bit less. Second, the interest she gives up on her own money, which could have stayed in the bank. Third, wear. Then take away the price rise she expects.

HOST 16:20

And the numbers for a cottage?

ECONOMIST 16:22

New numbers again. The bank and the loan charge 4 percent, wear is 2 percent, and buyers expect prices to rise 2 percent a year. That expected rise cancels the wear. What is left is the interest, made a bit cheaper by the tax rule: about 3 percent of the price. The cottage rents for 6,000 dollars a year, so it's worth about 197,000 dollars.

HOST 16:48

So which moves house prices more, easier credit or expectations? I'd bet on credit. If buyers only need their own money for 10 percent of the price instead of 20, prices should jump.

ECONOMIST 17:01

That's the tempting answer, but credit matters much less. Her own money and borrowed money cost about the same interest. Only the tax rule makes borrowing a little cheaper. So the price rises by only about 4 percent. Now suppose buyers expect prices to grow 1 point faster each year.

HOST 17:22

And what happens then?

ECONOMIST 17:24

The expected rise now more than cancels the wear, so the user cost falls from about 3 to about 2 percent. The price is the rent divided by the user cost. Divide by 2 instead of 3, and the price becomes about one and a half times as large. That's a rise of about 49 percent. And if rising prices make people expect more rises, the price keeps pushing itself up. That's how a housing bubble grows.

HOST 17:51

Last leaf: policy. The map lists corporate taxes, bubbles and housing. We've seen why bubbles are hard to spot. So start with taxes.

ECONOMIST 18:01

In 2017, the United States cut its federal corporate tax rate from 35 to 21 percent. In our model, that raises the capital firms want by about a third. One study found that firms with bigger tax cuts did invest more. But the cut cost a lot of tax revenue, so it is still debated.

HOST 18:22

And housing?

ECONOMIST 18:22

Rules that require bigger down payments barely move prices in the model. But they limit how much buyers borrow. And heavy borrowing can turn a fall in house prices into a banking crisis, as we saw in Chapter 10.

HOST 18:34

So in part three we saw that shares and houses are priced like machines: what they pay, divided by a user cost. And because that cost is small, expectations move prices a lot.

ECONOMIST 18:46

Exactly. Let's sum up.

HOST 18:50

Let's go back to our three goals. First, how does investment swing, and why?

ECONOMIST 18:56

It falls hard in recessions, and housing and inventories swing most. It swings because each year's investment is small compared with the capital firms own. And the capital they want depends on rates, taxes and expected prices.

HOST 19:11

Second, the user cost: interest plus wear, minus any expected price gain. A 35 percent tax raises it from 10 to 15.4 percent, and the firm wants about half the saws.

ECONOMIST 19:24

Right, because a saw's extra output rises only slowly as the firm owns fewer saws. Third, shares and houses are worth what they pay, divided by a small gap. So a small change in expected growth moves their prices a lot.

HOST 19:38

So what's the big message?

ECONOMIST 19:41

Every investment, in a machine, a share or a house, must earn what the bank pays. That one comparison sets the user cost of capital, how much firms invest, and what stocks and houses are worth. And because it rests on expectations of the future, investment swings sharply in booms and recessions.

HOST 20:03

Here's a question to think about. Islandia announces a tariff, a tax on imported goods. It will raise the price of power saws by 3 percent over the coming year, then stop. What happens to Harbor Boatworks' investment this year and next year, and why?

ECONOMIST 20:20

Pause the episode and think about it for a moment.

HOST 20:28

Okay, we're back. What's the answer?

ECONOMIST 20:30

This year, investment jumps. A saw bought now can be sold for 3 percent more later. That price gain lowers the user cost from 10 to 7 percent, so the firm wants about 1.7 times as many saws, and buys them now. Next year, the price gain is gone, and the user cost goes back up. Now the firm owns more saws than it wants. So it stops buying and lets wear shrink its capital. The tariff moves investment forward in time: a boom, then a bust.

HOST 21:02

Where can students try all this themselves?

ECONOMIST 21:05

Open the chapter's Colab notebook to change the interest rate and the tax yourself. The Model Apps have the same sliders, with no code. And the Data Apps show the investment and stock market data.

HOST 21:16

And next time?

ECONOMIST 21:17

In Chapter 18, we turn to the government. We ask: how much can a government borrow, and why does growth matter more than repayment?

HOST 21:27

That's it for today. Thanks for listening, and see you next time.