Low-Cost Leverage: The “Box Spread” Trade

December 9, 2021 – Last month, I published Part 49 of my Safe Withdrawal Rate Series, dealing with leverage in retirement. In that post, I surmised that the cheapest form of leverage likely comes in the form of a margin loan in an Interactive Brokers (IB) account. If you have the IB Pro account you have access to loan rates tied to the Federal Funds Rate plus a tiered spread ranging from 0.3% to 1.5%. Though, the really low rates don’t start until your loan reaches at least $3,000,000. For more manageable loan amounts that the average retail investor would use, we’re looking at a higher spread: 1.50% spread for the first $100,000 and 1.00% over the Fed Funds Rate for the next $900k. With the current effective Fed Funds Rate at around between 0.08% and 0.10%, that’s a very competitive rate. Certainly better than a Home Equity Line Of Credit (HELOC).

In the comments section, though, a reader brought up an idea for an even lower-cost method for borrowing against your assets: an exotic options trade called a “box spread”. I had heard of this trade before but never put much thought into it. And I certainly didn’t put any money into that idea. But just for fun, I researched this trade some more and even initiated one box spread trade on Monday, essentially issuing a synthetic $20,000 zero-coupon bond maturing in December 2026 at a very competitive interest rate, significantly lower what you’d get from IB.

So, in today’s post, I like to go through the basics of the Box Spread, how to implement it and how this trade could in fact give us a cheaper form of leverage than even the rock-bottom rates from IB. Let’s take a look at the details…

Box Spread Basics

A Box Spread involves four options trades, two long and two short option positions at two separate strikes. It can be designed to mimic a loan where we achieve a positive cash flow today and pay back the “loan” on the expiration date. Specifically:

  • Pick two strikes X1, X2, with X1
  • Sell one Call with strike X1
  • Buy one Call with strike X2
  • Sell one Put with strike X2
  • Buy one Put with strike X1
  • All four options have the same expiration date
  • All four options must be European Options, i.e., they are exercised only at the expiration, not before. This will eliminate single-name stock options because they are normally American Style with the option to exercise.
  • I’ve been told that this should only be done if you use “Portfolio Margin.” Not recommended for Reg-T. This trade is not allowed in cash accounts or IRAs.
  • One can now quickly confirm that the payout from this trade at expiration is precisely the spread between the strikes X2-X1, no matter where the underlying U ends up on the expiration date:

    • If U<X1 then the two calls expire worthless but the puts are both in the money. We pay out X2-U on the short put and we receive X1-U on the long put. Profit = X1-U-(X2-U) = -(X2-X1), i.e., we pay out X2-X1
    • If U>X2 then the two puts expire worthless but the calls are both in the money. We pay out U-X1 on the short call and receive U-X2 from the long call. Profit = U-X1-(U-X2) = -(X2-X1), i.e., we again pay out X2-X1
    • If X1≤U≤X2, i.e., the underlying finshies between the two strikes then the short Call and the short Put are in-the-money while the long options expire worthless. On the short Call we have to pay out U-X1 and on the short Put we pay X2-U, for a total loss of X2-U+U-X1=X2-X1.
    Payoff diagram of a 4600-4400 box spread.

    Because we have established a payoff of precisely -(X2-X1) regardless of the final underlying we have issued a synthetic zero-coupon bond for which we pay out X2-X1 (times the multiplier of the option, of course). So for example, I issued a box spread trade on the S&P 500 index with a 12/17/2026 expiration. The two strikes are X1=4400 and X2=4600. The multiplier is 100x. and the synthetic zero-coupon bond would have a notional of $20,000.

    If I have to pay back 100x$200=$20k in 2026, how much money did I get earlier this week on 12/6/2021? Well, financial markets are supposed to be efficient, so this implicit 5-year loan should trade at an implicit loan rate of roughly “the risk-free rate” over that time span. I put that phrase in quotation marks because there isn’t one single, generally accepted risk-free rate concept over a five-year horizon. But more on that later. Before that, I like to go through the different steps to initiate that trade on the IB trading platform…

    Implementing a Box Spread in an Interactive Brokers Account. A step-by-step guide

    Because four separate options are involved, it is crucial to keep the trading costs to a minimum. It would be prohibitively costly to initiate four separate trades for the four legs due to the large bid-ask spreads. But don’t despair… Interactive Brokers to the rescue! IB offers a Spread Trader where you can send packaged orders that give you simultaneous(!) trades for all four trades and only one single bid-ask spread. You still have to pay commissions for all four options, but that’s a drop in the bucket!

    Here’s my step-by-step guide:

    Step 1: open the TWS Spread Trader.

    Step 2: Select the underlying, i.e., the SPX index

    Step 3: Select either SMART or CBOE (I tried both and they both seem to get the job done)

    Step 4: Pick the spread trade: Box

    Step 5: In the menus on the left, we can select the parameters. Clicking while holding “Ctrl” or “Shift” will select multiple options. For example, I selected

    • The four expiration dates between December 2023 and December 2026
    • Strike 1 (which is the higher of the two strikes): I picked 4700 and 4800
    • Strike 2: I left that unchecked because the second spread is pinned down by the spread.
    • I set the point spread to 200 points, i.e., I target a $20,000 loan.
    • Trading Class: I set that to SPX (not really needed)

    Notice that in the beginning there must a gazillion of different box spread options. Think of the dozens of possible strikes, then all possible combinations of strikes, I guess N(N-1)/2 if my math is right, multiplied by all possible expiration dates. But once you cut down on the different possibilities by selecting your spread parameters, I’m left with 8 different box spreads: Two different strike pairs for four different expirations. Then click “Next”

    Step 6: To prepare for the next screen, where IB will display all the different spreads, we can pick how we want to arrange them. I like to create a matrix with the months in the columns and “Strike 1” in the rows. Then click “Finish”

    Step 7: And here we have the matrix of spreads. I right-click the 4700-Dec2023 spread. For some reason the Blue/Red fields for Buy and Sell are disabled. We have to enter the trade through the “Trade->Order Entry” menu option, see below:

    Step 8: That opens a new box. I enter a limit price of -199.0, i.e., a very low implicit interest rate for a 2-year loan. I know that this order will not be executed immediately. I can always play with the limit price later. And very important: You have to enter the limit price as a negative number. You are buying this at a negative price, so you will get a credit today and pay back the “loan” at the expiration date. Next, click the blue box “Buy”.

    Step 9: The order confirmation screen: Two things to make sure:

    1) The box spread is as we want: sell the Call at the lower strike, buy the call at the higher strike. Sell the Put at the higher strike, buy the put at the lower strike

    2) There is indeed a credit, i.e., we buy this spread at a negative(!) price. In this case, we receive a credit of $19,900 minus $5.89 for the commissions.

    I redacted my exact account values in the picture below. But notice that the box spread will have no impact on any of my margin requirements!

    Click “Transmit”

    Step 10: We can initiate a number of different spreads. For now, I just picked this one. If go to my “Pending” tab in IB I can now see the unfilled order, see below.

    Step 11: I can also copy and paste all the spreads from the “Pending” tab and put them into a designated tab on my trading screen. That’s what I did below. I had already done this for the 4600-4400, 4700-4500, and 4800-4600 spreads for the three far-out expiration dates: December 2023, all the way to December 2026. That’s useful because all trades on the “Pending Screen” disappear once you cancel the trade, But I like to monitor the quotes of all the different spreads without having to go through the Spread Trader mess every day. See my designated Box Spreads trading tab below. Also notice the one spread I already hold: 4600-4400 for 12/17/2026. It’s now worth -$18,705. In other words, part of the implicit interest I have to pay for the loan is already factored in as a loss in this position: -$310. Eventually, by 2026 the loss will be at around $1,600.

    Also, notice how massive the bid/ask spreads are. You never want to enter this as a market order! Only a carefully calibrated limit order! This brings me to the next question…

    Update 2/2/2022: The Finance Buff posted a similar article with the step-by-step process you can use at Fidelity.

    What price (and implicit interest rate) should we target?

    The -199 limit was just a dummy price that I knew will never get filled. Unless someone feels really charitable and wants to give me a 5-year loan for only about 0.1% APR. People will usually tie their implicit interest rate to the prevailing Treasury yields plus an appropriate spread. For example, when I initiated the trade on 12/6/2021, here were the Treasury yields according to Bloomberg:

    Treasury Yields as of 12/6/2021. Source: Bloomberg

    For December 2026, 5 years from now I want to get as close to 1.21% as possible. From what I gathered on the interwebs, a spread of 30-50bps (0.30-0.50%) seems appropriate, so I would target a 1.51-1.71% implicit interest rate. A useful tool for calibrating the target interest rate is the site www.boxtrades.com (thanks to reader Matt for bringing this to my attention) where you can see the log of past trades for SPX box spreads. December 2023 seems to be the most actively-traded box spread. The far-out options for 2024/25/26 were only recently added by the CBOE, so maybe they will garner some more action soon.

    Why is there a spread in the first place? Well, your counterparty, the lender in this context, would consider the loan not 100% free from default risk. Exchange-traded derivatives are extremely safe against default. Even if a counterparty cannot pay, the clearinghouse would guarantee the transactions. So, you still have the possibility of a clearinghouse going belly-up, but that’s extremely unlikely. But we still have to add a small spread on top of the Treasury yield to account for that unlikely event.

    In any case, for 5.03 years (don’t be sloppy, it’s a little bit more than 5 years: 5 years plus 11 days!), we’d target a box spread price of 200/1.0151^5.03 = about $185.50. And again, I have to input this as a -185.50 (minus!) limit price! At that limit price nothing happened. I kept increasing the limit price (i.e., make it less negative), and finally at $184.00 it was filled. But only after it sat for about 30 minutes. That’s an implicit interest rate of about 1.67% (actually 1.68% when I factor in the $5.89 commission), much more than I was willing to pay initially but still less than a 50bps max spread. Better luck next time. But then again, in the big scheme, even the 1.68% rate isn’t so bad. The AAA corporate spread is 0.58% and I borrowed at 0.47% above the 5Y-Treasury yield.

    If I try to gauge how much a margin loan with IB would have cost me, I can look at the expected path of Federal Funds Rates over the next 5 years, derived by the 30-Day Fed Funds Futures contracts. On 12/6/2021, the December rates for 2021 through 2026 were 0.10%, 0.67%, 1.30%, 1.70%, 1.86%, 2.07%. (the final number is the Nov 2026 contract, because the December 2026 contract is not trading yet). If we linearly interpolate over the five years, I’d get a 1.32% average effective Fed Funds Rate over this time span. Add to that a 1.5% spread for an IB margin loan under $100,000 and you’re looking at a 2.82% rate. Ouch! Even for a $1,000,000 loan where the spread is 1.5% for the first $100k and 1.0% for the next $900k for a weighted average 1.05% spread, we’d still look at a 2.37% expected margin loan rate at IB for 2021-2026. The box spread is decidedly more attractive than the margin loan option even at the most competitive broker I know of.

    What to do with my Box Spread Proceeds

    Well, I put in a limit order for the 4600-4400 box spread for December 17, 2026, without even thinking very hard about what I would do with the proceeds. And before I knew it, the trade went through at -184.00. So, I had $18,400 in my account that I didn’t really need. I felt a bit like that meme of a dog chasing cars. And suddenly the dog “catches” a car and realizes he has no use for a car. And no driver’s license and insurance either.

    Since I have enough cash flow from my options trading and interest and dividend income already and I have no use for a margin loan, I opted for putting the proceeds into a relatively safe investment. I purchased 800 Shares of the KeyCorp Preferred Shares, “I” shares. The ticker is “KEY PRI” on IB and “KEY-PI” on YahooFinance. KeyCorp is a regional bank in the Midwest with a $20b+ market cap. Solid business with stable earnings, and I already hold these exact Preferred Shares in my IB account. KEYPRI currently pays a fixed $1.53 annual dividend (paid quarterly), worth 6.125% of the $25 notional value. The shares are callable on 12/15/2026, just two days before the box spread loan comes due. Isn’t that convenient? If there were to be called (redeemed) at $25 apiece I’ll have exactly $20,000 available to pay off the box spread loan two days later.

    If the preferred shares are not called, they will then start paying a floating rate of 3.892% above the 3-month LIBOR rate, which is another widely used short-term interest rate. It’s not explicitly linked to the Fed Funds Rate, but usually very close, slightly above the Fed policy rate. In perpetuity. That’s a nice feature because nobody knows what the inflation and interest rate landscape will look like in 30+ years, so I hate locking in a fixed rate in a potentially perpetual preferred stock (nice alliteration, though!).

    One drawback of KEYPRI is that it is currently trading at above par, i.e., at a premium above the notional value. I bought the shares for between 29.79 and 29.80 on Monday. But that’s expected because the 6.125% yield is just way too high in today’s low-interest environment. If they are indeed called at $25 on 12/15/2017, I’d still make a decent 4.65% annualized IRR, 3.92% after taxes. Ideally, I’d hope for the shares to not be called and then just “float” after December 2026 with a pretty decent interest rate. Say, if the Libor rate is at a more sustainable long-term level of 2.75%, the shares would pay a dividend of 6.642% of the $25 notional. That’s pretty rich. If the price stays at $27, that’s still a 6.15% annualized yield. I’d probably not even sell at that price and let it run and rather come up with $20,000 from elsewhere to pay off the box spread loan. 6.15% minus 2% inflation, there’s a 4.15% safe withdrawal rate right there! Also, in that scenario where KEYPRI drops to $27, the IRR of the leveraged investment would be 12.46% pre-tax and 10.45% net of taxes. Pretty sweet. But of course, the absolute worst-case scenario would be the corporation going bankrupt and defaulting. Well, then I lost $20k. We will see how that all works out. I hope my blog is still around at that time and I can share the final tally with you!

    Update (12/9/2021 9:30PM): How will this impact your margin?

    Since people asked in the comments section: how will this impact my account margin? Let’s go through an example. Someone has $100,000 invested in stocks (or ETFs), $0 in cash, and gets a box spread “loan” which deposits $20,000 into the account. If the money is just sitting in cash, nothing happens to your margin. Your net account value is still the same (abstracting from the $5 in commissions, for simplicity). See below:

    Nothing really happens to your margin picture. You have a $100k account and you need to keep a minimum of $25k to keep your $100k in margin. That’s because the margin requirement for Portfolio Margin accounts in IB is 25% for stocks and ETFs. Of course, if your stocks go down by 50% then your margin cushion shrinks. But the same would have happened without the box spread trade.

    But of course, nobody would ever just issue a box spread “loan” and leave the money just sitting there unused. More likely would be one of the two situations: take the $20k and buy more stocks or withdraw the $20k and use them for something else (living expenses, down payment for a house, etc.). Or a combination of the two. Let’s simulate that in the table below. The top panel is for the box spread. The bottom panel is for the regular margin loan. The left side is for a $20k investment in stocks, the right side is for the $20k withdrawal.

    Absolutely, if you either buy more shares or withdraw the money it will shrink your margin cushion. If the market then tanks, it will worsen the situation. Also notice that from a margin perspective, the box spread loan and the margin loan are the same.

    Update 12/15/2021: What about the “Payments in Lieu of Dividends” issue?

    When you buy assets on margin you will likely lose some of the tax advantages of the assets you buy. Dividend income will then be reclassified as “Payment in Lieu of Dividends,” which will be taxed as ordinary income. That could pose a significant tax drag if qualified dividends, taxed at the same lower rate as long-term dividends, are now taxed like ordinary income. For us, this would raise the marginal tax rate from 15% to 22%. Even worse, municipal bond interest from our closed-end funds (such as NVG, NZF, NMZ) would go from tax-free to a 22% marginal tax. Ouch!

    It’s not clear how the broker will treat the box spread loan. You still have a positive cash position, but it’s not entirely clear if IB will factor the large short option position into the cash value. So far, so good though. I received a large dividend payment today on December 15 and it was all treated as “ordinary dividends”. Nothing was classified as the dreaded “payment in lieu of dividends.” Granted, most of the dividends were announced in November with a record date in late November when I didn’t even have any box spread loans. But one preferred share had a record date on December 13, after I initiated the box spread. Even that one is still listed as an ordinary dividend! I hope it stays this way for future dividends!


    Thanks to my reader “TIM H” who reminded me of this useful tool. In today’s post, I just wanted to give an update on what I learned and share my notes with you. A box spread is a neat tool to generate margin loans at rock-bottom rates. Here are the pros, but I also like to point out some cons.


    • With a box spread trade we can likely generate margin loans with a rate even lower than the already competitive IB rates.
    • While you can indeed write off margin interest, most people will not be able to do so effectively because you’d need to use itemized deducttions. With the massive standard deduction of almost $26k in 2022 (married filing jointly), you’d need a really large margin loan to get over that and/or a lot of other itemized deductions. And then, only the amount above the standard deduction is effectively useful as a deduction. On the other hand, the implicit interest you pay on the box spread loan is indeed fully tax deductible. The SPX options are Section 1256 contracts and the losses will be considered 60% long-term and 40% short-term. I generate substantial option trading profits with my put-selling and the margin loan costs will offset some of those gains. Even in the absense of any other capital gains you can write off up to $3,000 in capital losses annually against your ordinary income.
    • You can lock in a fixed interest rate over a specific time horizon.
    • The “loan” cannot be called. Unlike a margin loan, where your broker could just willy-nilly pull the rug from under you, the box spread is a fixed-term loan. Non-negotiable. If you use European-Style options like the CBOE SPX contracts I use. Never do this with American-style options on individual stocks or ETFs!


    • While the lock-in of the box spread implicit interest rate might be plus it can also be a headache. If the future path of the Fed Funds Rate is much lower than expected today, say due to another pandemic shutdown and recession, you might do better with the “variable” loan tied to the FFR. In today’s landscape, though, that’s hard to imagine because even if the Fed Funds Rate stayed at 0% for the entire 5-year period, you’d still pay 1.50%, not much lower than my box spread loan. But The Fed will have no choice but to raise rates in 2022 and 2023 due to the post-pandemic inflation shock.
    • Box Spreads are more cumbersome than simply drawing down a margin loan. It’s not for the faint-hearted! It involves some serious derivatives trading. In the example above, you trade four options with a total notional exposure of $100x(4400+4400+4600+4600) = $1,800,000. And all that get a $18,400 loan. But look at the upside: you can really impress people at your next cocktail party.
    • No room for error! Make sure you add the “-” sign in the limit order. A tiny fat-finger mistake and you might be out of a five-figure sum!
    • You’d need to file an additional tax form , if you don’t already trade Section 1256 contracts: IRS form 6781. But that’s not much of a burden. The neat thing about Section 1256 contracts is that you don’t have to itemize any of your trades. Just plug in one number, the net profts of your Section 1256 contract for that entire year and you’re good to go.


    PS1: Some purists might point out that the truly cheapest possible leverage would be to liquidate your stock and bond positions and then simply go long S&P and Treasury futures to keep the same stock and bond exposure as before. Keep the cash in short-term funds, like a money market fund or 3-Month T-Bills. And then slowly deplete that cash until you do the next swap of physical assets into cash + futures. You can now effectively “borrow” against your assets at the short-term risk-free rate because futures are indeed priced to replicate the return of the underlying minus the risk-free asset return.

    The advantage of this approach is that you might get the lowest possible “interest rate” but you also liquidate your assets over time with all the tax consequences. So, the estate planning option of letting the capital gains run until you can pass the asset on to your heirs and they get the step-up basis goes out the window. But if you don’t care about that and/or you reside in the 0% capital gains bracket anyway, this might be something to consider.

    Update 12/11/2021: Well, not so fast. It turns out that with a long futures contract you don’t get to borrow at exactly the risk-free rate either, but more likely at a rate risk-free plus about 0.3%-0.4% as folks pointed out in the comments section. If you factor in the 1.3% expected Fed Funds rate over the next 5 years and you add 30-40bps to that you end up right at the box spread implicit interest rate. There’s no free lunch.

    PS2: Another route suggested by a reader would be to buy a leveraged ETF. Say, for example, you need $50k to live on, then you’d sell $100k worth of assets and invest the remaining $50k in a 2x leveraged ETF.

    Not recommended! I looked at the returns of the three funds: IVV (S&P 500 index fund, i.e., 1x). SSO (S&P 500 2x) and SPXL (S&P 500 3x). Via Portfolio Visualizer, I got total return data for all three funds going back to 12/2008. Annualized standard deviations indeed scale just as expected. The leveraged ETFs have pretty much exactly 2x and 3x the standard deviation of the index fund, respectively. But the realized returns? Not so much! If I calculate the implied CAGR as

    Implied CAGR = Risk-Free + Leverage * (S&P500 – Risk-Free)

    …, i.e., what the theoretical futures non-arbitrage condition would imply, then the actual leveraged ETF returns lag behind that “theoretical number” very significantly. By 4% and 10% for the 2x and 3x leverage funds, respectively.

    S&P 500 index vs. Leveraged ETF performance. December 2008 to November 2021. Source: Portfolio Vizualizer

    Only a small part of this can be attributed to the expense ratios. The 2x and 3x funds are also subject to a lot of trading (churning?) and the dreaded “whipsaw risk” because every day the portfolio managers have to rebalance the futures positons to exactly match the target leverage. So, leveraged funds seem far too expensive when considering all the costs, not just the expense ratios! Not just in this context, but more broadly as well, I would stay away from leveraged funds!

    Thanks for stopping by today! Please leave your comments and suggestions in the section below! If you have traded box spreads before, please share your experience!

    226 thoughts on “Low-Cost Leverage: The “Box Spread” Trade

    1. Ever heard of writing an extremely OTM bull put spread as a box spread alternative? The idea is to sell the last put in the chain (e.g. SPX 7300) and buy one or two strikes down (e.g. 7200). You wind up with a vertical credit spread where the two option greeks are almost completely opposite and opposed leaving only Rho exposed — similar to a box spread.

      Stated another way, it’s a bull put spread born into existence just a few dollars away from max theoretical loss. There’s a teensy bit of net positive beta, positive delta, and positive gamma that only matters if the stock market rapidly goes vertical. If the market crashes those greeks go from rounding off to zero to actually zero and you lose roughly a case of beer.

      Why do this versus a box spread? Perhaps better fills? I don’t really know. But I’m curious if you’ve heard of it.

      1. Interesting idea. I get pretty good fills already with the box, but I can check this out. Obviously when you’re borrowing this sounds OK because if the market were to rally beyond 7300 points you owe nothing. But that positive risk is then priced into the vertical spread and might be quite expensive, probably at least another 50bps p.a. in implicit interest.

    Leave a Reply

    This site uses Akismet to reduce spam. Learn how your comment data is processed.