# At what interest rate should debt be used as a tool?

I know people love to preach the merits of being 100% debt-free, but when used intelligently, debt can be a powerful tool. All else equal, if I have debt at an interest rate of say 1% and have an instrument that returns 3%, that 2% gap is profit (I know for a variety of reasons (taxes, etc.) it's not that simple, but you get the point).

That said, it's usually not as simple as that. Many investment instruments carry a fair amount of uncertainty around their expected return, and the utility of a dollar lost might not equal the utility of a dollar gained. What guidelines do you use to determine the interest rate where it makes sense to keep the debt and invest rather than eliminating the debt?

• The problem with the borrow at 1%, invest at 3% and net 2% is that as an individual, it's difficult to do that in a way that is diversified enough to be ensured that you'll actually yield the return. Consider the business models of banks in the 1900's before Federal mortgage intervention and FDIC. Thousands of people would make deposits, hundreds of loans would be issued for mortgages, but uncertainty about the quality of the loans could cause a run on deposits, which would in turn cause the bank to fail. Commented Apr 27, 2011 at 12:11
• Don't forget risk when comparing interest rates. If your loan is at 5%, then paying that debt off is a guaranteed 5% return. If you don't discount your return based on risk, then you should go borrow money to play roulette (pays 35:1, that's 3600.0%). Commented Apr 27, 2011 at 17:35
• @Alex B While important, you are addressing two separate issues. Roulette does not have an expected return of 3600%; it has an expected return on the order of negative 5%. Roulette is stupid not because there is risk but because it has negative expected return. The stock market, on the other hand, might have an expected return of say 5% but the actual returns might be more or they might even be negative. Commented Apr 27, 2011 at 18:24

I've been taking all the cheap fixed-rate debt banks would like to give me lately.

What Rate?

In practice I find the only way I get a low-enough rate on a longish-term fixed-rate loan is to use collateral. That is, auto loans and home loans. I haven't seen any personal loans with a low enough fixed rate. (Student loans may be cheap enough if they're subsidized, I guess.)

Here's how I think of the rate:

If you look at https://personal.vanguard.com/us/insights/saving-investing/model-portfolio-allocations , the average annual return on 80% bonds / 20% stocks is 6.7%, with worst year -10.3%. That's a nominal return not a real return. If you subtract taxes, say your marginal rate (the rate you pay on your last dollar of income) is 28% federal plus 5% state, then if you have no tax deferral the 6.7% becomes about a 4.5% average, with reasonably wide variation year-by-year.

(You can mess with this, e.g. using tax-exempt bonds and tax-efficient stock funds, etc. which would be wise, but for deciding whether to take out debt, getting too detailed is false precision. The 6.7% number is only an average to begin with, not a guarantee.)

Say you pay 4.5% on a loan, and you keep your money in very conservative investments, that's probably at least going to break even if you give it some years. It certainly can and sometimes will fail to break even over some time periods, but the risk of outright catastrophe is low. If your annual loss is 10%, that sucks, but it should not ruin your life.

In practice, I got a home loan for close to 4.5% which is tax-deductible so a lower effective rate, and got an auto loan subsidized by the manufacturer for under 3%. Both are long-term fixed-rate loans with collateral. So I was happy to borrow this money paying about a 3% effective rate in both cases, well below my rough threshold of 4.5%.

I do not, however, run a credit card balance; even though one of my cards is only 7% right now, 7% is too high, and it's a floating rate that could rise. The personal loans I've seen have too-high rates also.

Thoughts

• One of the reasons I take the loans is liquidity. If I have \$N in cash and then owe relatively small payments over 6 years, then that is a lot more flexible than not having cash and owning a car. If not having cash and owning the car outright ever becomes desirable, then pay off the loan, and be in that situation overnight. The other direction is often impossible.
• Another reason I take the loans is that it makes budgeting easier. I like to think of major assets (house, car, etc.) in terms of monthly cost. Then you can see how much more you need per year (and how much more you have to save for retirement) if you spend a certain amount on the car. Of course if you pay cash you can still compute the monthly cost, but it's just a little easier if it's actually a monthly cost, in my opinion.
• If you need a credit history, getting a loan can give you one, though a credit card is probably a simpler way.
• I have a fair amount of "wiggle room" (solid emergency fund, no trouble covering the debt payments). The more trouble you'll have with the payments, the less you should be buying anything at all probably, let alone with debt.
• You probably shouldn't be considering your emergency fund in this calculation. That is, don't consider using the emergency fund instead of getting the loan, and don't consider putting the emergency fund in anything higher yield than cash.
• Borrowing money at a low rate puts you in a good place if interest rates rise (or, probably saying the same thing, if inflation increases).
• Borrowing money (even at a low rate) would be bad if we get deflation. Thankfully that hasn't happened in general since the Depression, but it is happening now in housing specifically, and I'm sure you've seen on the news how it's bad for those who borrowed money to buy.
• Borrowing money at a floating rate is a risky idea. You lose in both inflation and deflation scenarios.
• I'm not sure the pay-cash-or-get-the-loan decision is all that important in the scheme of things. It isn't likely to be make-or-break on your overall financial situation. So you may as well do what makes you feel comfortable. For me, extra flexibility makes me comfortable, but for others, debt makes them uncomfortable. Both are valid.
• My discussion here assumes you have the option to pay cash. If you couldn't pay cash, you have nothing to invest, and no cash to keep in order to increase flexibility.

Overall I think using debt as a tool requires that you're already financially stable, such that the debt isn't creating a risky situation. The debt should be used to increase liquidity and flexibility and perhaps boost investment returns a bit.

Where you're likely to get into trouble is using debt to increase your purchasing power, especially if you use debt to buy things that aren't necessary.

For me the primary reason to use debt is flexibility and liquidity, and the secondary "bonus" reason is a possible spread between the debt rate and investment returns.

• What if you had some significant net worth: all your assets (e.g. car) paid off and a few \$100K of your own in the bank. If you're already liquid in that way, would you still borrow money when you buy a house or car, just so that you'd have even more money to invest? Commented Apr 27, 2011 at 2:46
• yes, I'm not rich but I could pay cash for a sensible car. I don't like "tying up" money in the car that way; I'd rather be able to leave the money invested where it is, and have the option to use it in emergencies too if it came to that. Also a subsidized sub-3% loan costs the manufacturer money and I feel like I'm leaving a discount on the table not taking it. I'm not sure the amount of net worth really changes the calculation once you save enough that paying cash is an option. Commented Apr 27, 2011 at 2:59
• An example in the corporate world, often companies will have a bunch of cash and then also issue bonds. The reason they do this is that the bonds have a low rate (it's probably during a boom period). Then they have even more cash around. At a later time, maybe they decide to buy another company, or maybe there's a recession, and at that point it might be hard to issue new bonds. Fortunately, they already borrowed money when it was cheap, and now they have it when they need it. Same idea for individuals; borrow when it's cheap, not when you have to. Ideally, you never have to borrow. Commented Apr 27, 2011 at 3:39
• The problem with those "subsidized 3% car loans" is that car prices are negotiatable, and that subsidy is paid from the same negotiating margins. I.e. by declining the subsidy, you can negotiate a lower cash price. This is generally sensible; many car dealers and manufacturers would prefer cash now. Commented Apr 29, 2011 at 8:53
• yeah, in the past I've bought a car and explicitly had the discount or low rate choice, and you could compute the present value of the options to decide. This time though I don't think the dealer got money if you skipped the cheap loan so it was separate. Based on lots of shopping and price reports. But who knows really. Commented Apr 29, 2011 at 12:07

Money is a commodity like any other, and loans are a way to "buy" money. Like any other financial decision, you need to weigh the costs against the benefits.

To me, I'm happy to take advantage of a 0% for six months or a modest 5-6% rate to make "capital" purchases of stuff, especially for major purchases. For example, I took out a 5.5% loan to put a roof on my home a few years ago, although I had the money to make the purchase.

Why did I borrow? Selling assets to buy the roof would require me to sell investments, pay taxes and spend a bunch of time computing them.

I don't believe in borrowing money to invest, as I don't have enough borrowing capacity for it to me worth the risk. Feels too much like gambling vs. investing from my point of view.

• Although there are wildly different social connotations, investing really is gambling. However, there is a difference between gambling with negative expectation (e.g. almost everything a casino offers) and gambling with a positive expectation (e.g. a skilled card counter or stock market index fund). Commented Apr 27, 2011 at 12:57
• I know it's a little pedantic, but I don't think there's much difference between borrowing in order to avoid selling investments, and borrowing in order to buy investments. In both cases you end up in the same place; with a loan and with investments that in theory you could sell to pay off the loan. To me the important distinction is between borrowing that's relatively safe, because e.g. it's backed by collateral alone, long-term and fixed rate; and say margin debt. Margin debt is a floating-rate loan that the brokerage can revoke at any time and will revoke at inopportune times. Commented Apr 27, 2011 at 15:00
• IMO, you need to internally compute the expected return against the potential for loss. Paying for money to invest creates a bias to seek short-term versus long-term gain. Riding out a bad market requires a mental tenacity beyond what I have when you're paying interest for the privilege of losing money. That's why I see that practice as a gambling activity vs. an investment. Commented Apr 27, 2011 at 20:22

Average return on the S&P500 over the last 10 years has been 1.6 %; so if you'd invested in that with money borrowed at 3 % you would have lost (so far).

Investing with borrowed money implies you think you can beat the market: that you're a cleverer investor than whoever decided to lend you the money.

Whoever decided to lend you the money decided that you are the best (return/risk ratio) investment for their money.

It might make sense to invest borrowed money if you don't need to pay it back if things go wrong: if you're an investment professional whose bonus depends on the profit you make, but who won't need to repay any loss.

It might also makes sense to borrow money if you're going to 'add value', e.g. sweat equity: for example if you use it to renovate a house or (if you're a business) to hire more staff. But the question was "What guidelines do you use" and the answer is, "I don't make passive investments with borrowed money."

My Dad did it, i.e. didn't repay his mortgage as soon as he could have: but that was because (back in the '70s) he had a long-term (government-sponsored) mortgage for about 1.5 % (designed to help first-time buyers or something like that), at a time when banks were paying higher interest rates on (ultra-safe) deposits.

• While you make an interesting point, I'm not exclusively talking about money that was borrowed purely with the intent to invest. Conceivably someone could have a home, auto, or student loan with a sufficiently low interest rate as to make it desirable to invest rather than paying off additional principle on the loan. Commented Apr 27, 2011 at 11:54
• @Michael - Same thing: if e.g. I have a mortgage at some 'x' % and make some extra money from my job, then I should invest that money (instead of paying down my mortgage) if-and-only-if I'm willing to bet that I <del>can</del> will get a better than an x % return on my investment (which, in fact, I'm probably not willing, for any significant value of x). Commented Apr 27, 2011 at 12:08
• The note I'd add here is that investing in the S&P500 alone probably isn't the way to go, or if you are doing that, you have to look at 20-30 year returns, not the last 10 years. A portfolio with only 20% stocks and 80% investment-grade bonds is probably good enough to grow 4.5% (pre-inflation return) on average. blog.ometer.com/2010/11/10/… goes into more detail on why I dislike 100%-stocks portfolios. Over the last ten years, a balanced portfolio did fine: quote.morningstar.com/fund/… Commented Apr 27, 2011 at 15:10
• btw, I don't agree that investing with borrowed money implies you think you can beat the market, because the lender doesn't have your same return/risk profile or opportunities. For example, the lender has no way to "collect" the tax subsidy on home mortgage rates, but you do. Similarly, a car manufacturer has reason to subsidize a loan and give you a below-market rate, which is just a way to put their cars on sale - their motives aren't all about investment returns. For leverage in general, modern portfolio theory says varying leverage (0 to over 100%) is the best way to boost risk&returns. Commented Apr 27, 2011 at 15:14
• "Average return on the S&P500 over the last 10 years has been 1.6 %; so if you'd invested in that with money borrowed at 3 % you would have lost (so far)." Huh? OP said he hypothetically borrows at 1%. The worst decade in a long time and he'd have come out ahead by your numbers. In normal times, he'd have quite the profit. Commented May 6, 2011 at 17:19

This is a very interesting question. I'm going to attempt to answer it.

1. Use debt to leverage investment. Historically, stock markets have returned 10% p.a., so today when interest rates are very low, and depending on which country you live in, you could theoretically borrow money at a very low interest rate and earn 10% p.a., pocketing the difference. This can be done through an ETF, mutual funds and other investment instruments.

2. Make sure you have enough cash flow to cover the interest payments! Similar to the concept of acid ratio for companies, you should have slightly more than enough liquid funds to meet the monthly payments.

3. Naturally, this strategy only works when interest rates are low. After that, you'll have to think of other ideas. However, IMO the Fed seems to be heading towards QE3 so we might be seeing a prolonged period of low interest rates, so borrowing seems like a sensible option now. Since the movements of interest rates are political in nature, monitoring this should be quite simple.

4. It depends on you. Since interest rates are the opportunity cost of spending money, the lower the interest rates, the lower the opportunity costs of using money now and repaying it later. Interest rates are a market mechanism so that people who prefer to spend later can lend to people who prefer to spend now for the price of interest.

*Disclaimer: Historically stocks have returned 10% p.a., but that doesn't mean this trend will continue indefinitely as we have seen fixed income outperform stocks in the recent past.

• Do you have a source for your "10% p.a." claim? I'm finding that a bit hard to believe compared to the other numbers I've seen quoted. Commented Apr 27, 2011 at 7:02
• personal.vanguard.com/us/insights/saving-investing/… supports 10% average, but caveats are: it's more like 6-7% after inflation, and you need a 30 year time horizon to "count on" being in this ballpark. Also you have to be a little flexible about the end of the 30 years i.e. not have a forced cash out right on schedule which may be during a recession. Oh, and subtract taxes too. Commented Apr 27, 2011 at 15:30

It's not so much the rate of the debt as it is the total cost of the debt relative to the gain you expect to see from using it to purchase something of value.

I've known people who were quite happy to pay 12% on personal loans used to buy investment properties for flipping. They're happy to pay that because conventional loans from banks require too much documentation and out-of-pocket expense. For some investors, 12% without all of the documentation burden is money well spent.

So if I'm the investor, and the interest on this 12% loan is \$5,000 and I can flip a property for \$20,000 after all of the other expenses, then the 12% loan was an enabler to netting \$15,000 profit.

This post has a great discussion on the topic.

Basically, there is no single interest rate above which you should pay off and below which you should keep. You have to keep in mind factors such as

• Whether or not the interest for that debt is tax deductable.
• What is the expected return of the alternative investment option.
• Obviously the taxes matter. I thought that estimating expected returns of widely available investment options was implicit in my question. For instance, one might quote historical returns of the stock market, but one could argue that using those historical results is a bad idea. I'd like to see such issues addressed. Commented Apr 27, 2011 at 1:58

It's tough to borrow fixed and invest risk free. That said, there are still some interesting investment opportunities. A 4% loan will cost you 3% or less after tax, and the DVY (Dow high yielders) is at 3.36% but at a 15% favored rate, you net 2.76% if my math is right. So for .5%, you get the fruits of the potential rise in dividends as well as any cap gains. Is this failsafe? No. But I believe that long term, say 10 years or more, the risk is minimal.

I don't really see it as worth it at any level because of the risk.

If you take \$10,000,000 using the ratios you gave making 2% return. That is a profit of \$200,000. Definitely not worth it, but lets go to 20% profit that is \$2,000,000. To me the risk involved at beint 10 million in debt isn't worth it to make \$2,000,000 quickly it would be pretty easy doing something wrong to wipe out everything.

• I like how I get a downvote on an opinion question. heh. Commented Apr 27, 2011 at 1:43
• You can't really make blanket statements about it not being worth the risk without even considering what the risk is...especially since those were made up numbers for demonstration purposes. Furthermore, there's no need to take this to some ridiculous extreme...what would be valid to do with say \$10,000 might make less sense with \$10,000,000. It's rather odd that you jumped to those sums, which are clearly out of your (and post people's) comfort range instead of considering more realistic numbers. Commented Apr 27, 2011 at 1:46
• That was the point of my answer, and the reason I figured I was down voted. Even though you tried to take risk out of the equation in the question you still include it in an opinion of an answer. The premise of your question is to try to be numbers only, but you dislike big numbers in an answer. I personally wouldn't feel comfortable with \$10,000 in debt to make \$2,000 and Im intelligent with money. I am working my way out of debt and the hassle of dealing with debt isn't worth it to make it a powerful tool. Commented Apr 27, 2011 at 1:57
• I'm not trying to take risk out of the equation at all...that is one of the most important factors in coming up with an answer. You didn't really address it though...you just blindly assumed that a hypothetical situation was risky rather than trying to come up with any assumptions or estimates around quantifying that risk. You seem to be punting on the issue by saying that if risk > 0 then nothing else matters. Commented Apr 27, 2011 at 2:04
• Going back and reading a few more times your question I can see how you can "include" risk in it. However, I am meeting more and more people that are starting to follow the risk > 0 is there so nothing else matters. Yes, I did jump into a hypothetical situation because I can't quantify the risk, anymore. So at the end of the day the numbers don't matter. Commented Apr 27, 2011 at 2:09