Executive Summary
Clients who need to improve their prospects for retirement generally have three options: spend less, save more, or retire later. Technically, there is a 4th option - grow faster - but it is typically dismissed due to the risk involved in investing for a higher return. In practice, clients rarely seem to dial up the portfolio risk trying to bridge a financial shortfall in retirement, and taking out a margin loan just to leverage the portfolio to achieve retirement success would most assuredly be deemed imprudent and excessively risky. Yet at the same time, a common recommendation for accumulators trying to bridge the gap is to keep any existing mortgages in place as long as possible, directing available cash flow to the investment portfolio, and giving the client the opportunity to earn the "risk arbitrage" return between the growth on investments and the cost of mortgage interest. There's just one problem: from the perspective of the client's balance sheet, buying stocks on margin and buying stocks "on mortgage" represent the same risk and the same leverage, even though our advice differs. Are we giving advice that contradicts ourselves?
The inspiration for today's blog post comes from several conversations I've had recently with planners since my blog post a few weeks ago about how planners never tell clients to take out a loan to contribute to a retirement plan, yet willingly tell clients to contribute to 401(k) plans instead of paying down a mortgage, despite the fact that it's comparable when looking at the entire balance sheet. In the discussions both in the comment section of the blog and offline, I began to realize we as planners have some oddly contrasting advice about debt and what is and is not risky.
For example, imagine 3 clients, who we'll call A, B, and C. Client A has a $500,000 residence with a $500,000 fixed-rate 30-year mortgage at 5.0%, and a $1,000,000 conservative portfolio with a beta of 0.5. Client B has a $500,000 residence with no mortgage, and a $1,000,000 conservative portfolio with a beta of 0.5 that is collateral for a $500,000 variable-rate margin loan with a 5.0% interest rate. Client C has a $500,000 residence with no mortgage and a $500,000 aggressive all-equity portfolio with a beta of 1.0.
If you asked a typical planner which of these clients was the least risky, virtually all of them would answer client A, with the stable mortgage and the conservative 0.5 beta portfolio. After all, client B has a giant margin loan against his portfolio - "clearly" risky - and client C is high-risk due to the incredibly volatile portfolio being fully invested in equities with a beta of 1.0, twice that of client A.
The problem is, when you look at the impact of short-term market volatility on the client's entire balance sheet, client A is not less risk-exposed than client B. In fact, all three clients have the exact same exposure to market volatility! To see why, let's look first at the current balance sheet for all three clients, as shown below:
|
Client A |
Client B |
Client C |
||||
|
Assets |
Liabilities |
Assets |
Liabilities |
Assets |
Liabilities |
|
|
Residence |
$500,000 |
$500,000 |
$500,000 |
|||
|
Portfolio |
$1,000,000 |
$1,000,000 |
$500,000 |
|||
|
Loan |
($500,000) |
($500,000) |
$0 |
|||
|
Net Worth |
$1,000,000 |
$1,000,000 |
$1,000,000 |
|||
As the balance sheet shows, all three clients have the exact same net worth. In addition, while Client C has a portfolio half the size of clients A and B, the portfolio has twice the volatility. Consequently, if the markets rally 10% in the near term, all three clients experience a comparable result; clients A and B grow their portfolio by 5% (since their beta is 0.5) on a portfolio of $1,000,000 and client C's portfolio enjoys the full 10% return on $500,000. Consequently, after the rally, all three clients have a net worth of $1,050,000, as shown below (a similar $50,000 loss occurs for all three clients with a 10% market decline):
|
Client A |
Client B |
Client C |
||||
|
Assets |
Liabilities |
Assets |
Liabilities |
Assets |
Liabilities |
|
|
Residence |
$500,000 |
$500,000 |
$500,000 |
|||
|
Portfolio |
$1,050,000 |
$1,050,000 |
$550,000 |
|||
|
Loan |
($500,000) |
($500,000) |
$0 |
|||
|
Net Worth |
$1,050,000 |
$1,050,000 |
$1,050,000 |
|||
In fact, any level of portfolio returns will be expressed evenly across all three of the client balance sheets, because their effective exposure is the same - client C achieves a beta of 1.0 by direct investment, and clients A and B achieve the same result by having a portfolio with half the beta and leveraging it to twice the size.
However, there are two important caveats to the scenario. The first is that in very extreme market declines, the outcomes are no longer even. While client C is limited in a market decline to losing his whole portfolio and nothing more, clients A and B could theoretically lose so much in the portfolio that there isn't enough left to pay back the loan entirely. At that point, clients A and B can actually experience worse losses than client C; they can actually forfeit a portion of the equity in their residence to pay back the value of the loan if the portfolio falls below $500,000. Leverage has additional risk.
The second caveat is that at the end of an entire year of this process, client C is also better off - because client C doesn't bear the cost of debt service. If we assume that the 10% market return occurs not in a short period of time, but instead over the span of a year, then at the end of the year client C has a full net worth of $1,050,000. However, client A must reduce his net worth by 5.0% x $500,000 = $25,000 of mortgage interest, and client B similarly loses 5.0% x $500,000 = $25,000 of margin loan interest. Thus, at the end of the year, client C is actually the wealthiest at $1,050,000; the other two clients only have $1,025,000! (If we assume that clients A and B pay the mortgage interest from cash flow, the result is the same, as client C would have had $25,000 of free cash flow to save and contribute to his account, producing the equivalent difference.)
In other words, "just" owning a more volatile (i.e., "riskier") portfolio is actually the least risky way to dial up the return, superior to both buying stocks on margin or "on mortgage"! It creates greater wealth at the end of the year by saving loan interest, and eliminates the risk of a negative equity situation where the outstanding loan could exceed the value of the investment account, digging into the equity of the residence. Accordingly, this means that even if clients A and B already had a relatively aggressive portfolio, the best path for client C is still to just own even more volatile stuff in the portfolio itself; when Clients A and B have a beta of 0.5 then Client C has a beta of 1.0, but if clients A and B already have a beta of 1.0, then client C should buy small caps, emerging markets, 2x leveraged ETFs, and/or whatever else is necessary to dial up the beta to 2.0! It turns out that is still a less risky, and more cost-efficient, way to get to the retirement goal than simultaneously keeping a mortgage and an investment portfolio (or keeping a margin loan and an investment portfolio) when looking at the client's entire balance sheet! (Of course, if the size of the loan relative to the portfolio was smaller and entailed less leverage, then client C's portfolio wouldn't have to be that much more volatile than the portfolios of clients A and B.)
Secondarily, it is striking that the scenario for the mortgage loan - typically viewed as a conservative route of using "good" debt - and a margin loan - typically viewed as a risky route of using "bad" debt - yield substantively identical results. For large margin loans, the interest rates are competitive, and it's even true that both can potentially be deducted for tax purposes (as mortgage interest or investment interest, respectively) and both often utilize variable interest rates (although mortgage interest at least can be fixed). Functionally, the only real difference between the two happens to be the collateral involved; yet it's not entirely clear offhand why buying stocks using stocks as collateral is "risky", but buying stocks and using your home as collateral is less risky!?
Thus, it may be somewhat surprising that relative to conventional wisdom, the most prudent route to retirement success - for those who do want/need to dial up the return somehow - is to pay down the mortgage (or margin loan) out of the portfolio and dial up the volatility of the portfolio itself, rather than trying to earn the "risk arbitrage" between the portfolio return and the loan interest rate. Viewing the portfolio on its own, the former may "feel" more risky, but relative to the client's entire balance sheet, it's actually less risky, as it eliminates cash flow requirements, the danger of negative equity, and accrues wealth faster by also eliminating the real cost of interest. Even if that means buying a boatload of the "riskiest" high-beta stuff you can find. If the client isn't comfortable with that level of risk, then you certainly shouldn't invest that way - but that also means you probably shouldn't be duplicating the effect by buying stocks on margin or "on mortgage" by keeping a giant mortgage outstanding while investing in the portfolio at the same time.
As some critics point out, this approach could potentially be very challenging to clients who wouldn't have much of any portfolio left over if they paid off their mortgage - for instance, if it was a $1,000,000 residence with a $400,000 mortgage, and a $400,000 portfolio on the side. "Surely," they say, "it can't be right for a client to pay off the mortgage in this scenario; he'd had no portfolio left at all!" To which I can only reply: "If the client's concern is that he'd be left with $1,000,000 of equity in the home and no equity in the portfolio, the problem isn't the size of the mortgage, it's the amount of assets the client chose to invest in real estate!" In other words, if the client doesn't want all of his net worth tied up in a $1,000,000 piece of real estate that makes him "house rich and cash poor"... then the solution is not to borrow the equity back out via leverage, the solution is don't buy such a huge real estate asset in the first place! (And if you do, keep a home equity line of credit open so you can borrow if there's a real emergency that merits the borrowing risk, not as a way to leverage your portfolio!)
In the end, though, the fundamental point remains: from the perspective of the client's entire financial balance sheet, buying stocks "on mortgage" is remarkably comparable to the risk of buying stocks on margin, which is almost (but not quite!) as risky as just investing in a portfolio that is twice as volatile in the first place. And while the last option may feel the scariest, it turns out that it's actually the most conservative, and least expensive, route. If that's not a comfortable level of risk in the portfolio, then a giant mortgage (or margin loan) should be a concern, too. In fact, I suspect many clients intuitively understand the implicit risk involved with mortgage debt and leverage - which is why so many people intuitively have a drive to pay down their mortgages quickly. And if the risk level isn't comfortable, that's actually the exact correct prescription for them, too.
So what do you think? Have you ever compared a mortgage loan to a margin loan? Or a mortgage loan to a portfolio that's twice as volatile? Are we or our clients compartmentalizing risk by separating the loan from the investments? Is it appropriate to do so? Or should it be viewed from the perspective of the client's entire balance sheet? Should we be telling clients to buy fewer stocks "on mortgage" and instead just invest in more volatile portfolios directly (and if they don't want the latter, don't let them do the former, either)?
Really interesting analysis, Michael. I certainly have observed that it’s almost standard to recommend keeping a mortgage both for tax write-off and to free up money for investments, with an almost off-hand disclaimer about the risk that the interest-rate arbitrage isn’t guaranteed to pan out. Equating investing “on mortgage” with investing on margin is a delightfully concise way of making someone understand the risk of doing this very common (and I think probably not well-enough understood) thing!
This is a great post on an interesting topic. However, I think it would be constructive to add a complication to the discussion. For individuals who are choosing between paying down a mortgage ahead of schedule and contributing to a retirement plan there are the added elements of tax deferral and employer matching. Given a fixed rate mortgage in today’s low interest rate environment and a retirement plan with decent matching I would think the comparison would look quite different.
Collin,
Certainly, a match from an employer for a retirement contribution represents an exception to the general framing here. Well, not an exception, but a guaranteed instantaneous 100% return would certainly make the risk worth it.
That still only explains the value of making a contribution up to a 401(k) match, though, and not for additional contributions or savings, nor to why you would keep a taxable investment account while still maintaining a mortgage.
As for the tax deduction, it may partially reduce the amount of leveraged risk involved, but there’s still leverage and risk to this. I may do a follow up on the tax deduction later in particular, as I’m seeing a lot of confusion about what its value is and is not.
Thanks for the feedback!
Respectfully,
– Michael
Michael: How do we chunk this down for a client? Most of my clients will have a hard time absorbing this paradigm and I’m pondering how I deliver the message. My plans usually have the client paying off their mortgage just short of retirement and I use http://www.whatsthecost.com to illustrate aggressive mortgage payment and what it will mean in reduced years. (I also use it for snowballing debts.) I motivate them by showing them their reduced cash flow needs in retirement and how much capital they would need to generate a mortgage payment over x years. I like your risk comparison approach, just wondering how to deliver it.
Michael,
I must take issue with a number of your positions:
*1) “In other words, “just” owning a more volatile (i.e., “riskier”) portfolio is actually the least risky way to dial up the return, superior to both buying stocks on margin or “on mortgage”!*
I’m surprised that you used an overly simplistic scenario of returns to come to a conclusion that portfolio C is in fact less risky than portfolio A or B. What happens to your theory when you populate it with the experience of the last three calendar years? Let’s assume, a bit more realistically, that the .5 beta portfolios are invested 50/50 in the S&P 500 and the BarCap Agg (rebalanced annually) and the 1.0 beta portfolio invests in the S&P 500. We’ll assume the mortgage is interest only @ 5.0% with a marginal tax bracket of 28%, ignoring state income tax implications – (total after-tax interest cost roughly $54,000). Which scenario comes out ahead, net of any liability? Portfolio A or B, by about $70,000. Your example ignores the returns of the 0 beta assets and the implications of deep negative returns, which, seemingly, would be THE motivation for warning readers of using mortgage leverage for portfolios.
*2) “Functionally, the only real difference between [mortgage / margin] happens to be the collateral involved; yet it’s not entirely clear offhand why buying stocks using stocks as collateral is “risky”, but buying stocks and using your home as collateral is less risky!?”*
By far the most common mortgage instrument is the 30 year fixed rate loan. I believe it’s the most reasonable benchmark to use as a comparison to margin loans. The most obvious reason that a margin loan would be more risky is the variable rate – which I believe you acknowledged at some point. If interest rates increase to the point where negative carry is introduced at precisely the wrong time (read asset prices falling), the investor could be forced to sell into distressed prices. No risk of rising rates for a 30 year fixed rate mortgage.
Second, I have never heard of a margin call for a mortgage. But for a margin loan, if your portfolio balance falls too far, you could face a capital call or, again, be forced to sell into depressed prices to reduce your margin balance. To me, it seems that margin loans are a more risky and expensive (in the long-term) form of leverage; and we didn’t even address the possibility that the mortgage is non-recourse which would only offer the borrow more protection in the form of an effective put option on their house! Moral obligations ignored.
I’m not suggesting that obtaining a mortgage to invest in securities is a good idea. But I dare not say that doing so could never make sense. Today’s interest rates present some interesting options for someone approaching retirement with home equity, a fixed pension and the willingness to take on a mortgage, for example. I am unwilling to say that I would approach that discussion with a closed mind. Said another way – over the next 30 years it’s hard for me to believe that a balanced portfolio wouldn’t earn over 3% annually (4% interest / 25% marginal tax bracket). Obvious conflict of interest acknowledged.
Michael,
The mortgage vs margin question has received fair attention. I agree that mortgages are safer. It’s a matter of matching assets to liabilities. Long-term assets like stocks are better matched against long-term liabilities such as a mortgage.
What I think requires much more careful consideration and caution is the suggestion of shifting leverage to the investment portfolio through 2x strategies or dialing up beta. Earlier comments referenced how going from a 50/50 portfolio to 100% equities would have failed as a strategy recently. Failure for these purposes should mean it didn’t produce 2x the returns (positive or negative of the 50/50) Going forward there is certainly no assurance that higher beta equals 2x relative returns. In regards to 2x strategies, these are not only risky but deceiving. You mentioned going 2x the S&P and 2x the Barcap agg. Such a strategy does not guarantee to double the return of a 50/50 portfolio. Run the numbers on what would happen if the S&P drops 30% year 1 and then fully recovers in year 2. The 2x strategy is not made whole. Or consider that 2x ETFs are intended to track only the daily performance on an index, and are almost guaranteed to have considerable tracking error over an extended period of time. Leveraged ETFs can fail horribly at achieving the intended goal of 2x an indices’ return over time. Other strategies that target higher beta by using assets like small caps, emerging markets, or real estate are similarly failed. Beta is a measure of relative price volatility. Beta is not going to be a reliable method of achieving a long-term ratio of relative performance.
Anthony,
Indeed, I realize there are tracking error problems over time with the actual 2X leveraged funds out there in the marketplace – that alone could be covered in a multi-article series, so I can’t give it due justice in a single blog post where it isn’t even the primary point of the article.
But honestly, I still have an incredibly difficult accepting “the reason why you should borrow money against your house to invest is because all other ways of making your portfolio more volatile are too unreliable” – per the criticism of 2X funds, diversification, etc.
Is our point to the general public REALLY meant to be “you must use your house as leverage to achieve your goals, because it’s the best way to make your balance sheet twice as volatile without having tracking error and we can’t come up with any other way to take on more risk in pursuit of higher returns”!? At least with EVERY SINGLE investment strategy you criticized, I can’t actually end out with NEGATIVE equity the way I can by using debt!
Are we really unwilling to entertain ANY other way to achieve someone’s long-term growth goals besides extracting huge exponential leverage by borrowing money with our home as collateral?
How did we get to the point where “borrowing money against the residence” is the “safest” strategy by which all other investment strategies are benchmarked?
Respectfully,
– Michael
Anthony,
Viewed another way:
If we want to have a conversation about the best way to construct a volatile investment portfolio, then let’s have that discussion as a profession.
But I still fail to see why a volatile portfolio – where the worst case scenario is just that the portfolio declines in value – is BETTER than a scenario that can result in bankruptcy and eviction.
I don’t care how volatile you make the portfolio, if you don’t have LEVERAGE you can’t go NEGATIVE. That’s not true when you borrow against the home.
Do we really think it’s prudent to risk bankruptcy and eviction because we can’t think of any other way to make a portfolio more volatile?
Respectfully,
– Michael
I think the discussion is skewed because we are dealing with extremes, such as attempting to double client investment returns with 2x strategies, whatever they might be.
Our industry is sufficiently clever in ways to make portfolios more volatile, just not in obtaining a targeted level of returns. If an advisor used your strategy, it would be best to constrain their target return multiple substantially.
Maintaining a mortgage involves known costs and risks. These risks should not be discounted, and understanding that the mortgage creates leverage should be made clear. You make a good point that this attribute of mortgages is not always appreciated, and the true impact on the overall balance sheet not always observed.
We should never tell someone they MUST use a mortgage as a tool for arbitrage.
Regards,
Anthony
Anthony,
But the point remains – if the client IS trying to leverage up the returns on their balance sheet by maintaining a giant mortgage, WHY is that the “only” acceptable manner to generate magnified returns?
Why is excessive leverage on a “conservative” portfolio safe, but no leverage on an aggressive portfolio unreasonable?
Because it’s worth noting, virtually every one of the great investment disasters in history, from Long Term Capital Management to the financial crisis of 2008, were built on the back of excessive leverage with investments that were “thought” to be conservative until it turns out they weren’t.
Why do we continue to pretend that a conservative investment with leverage is so much less risky than a volatile investment with no leverage – even/especially when they’re targeted to generate approximately comparable returns – despite the clear evidence to the contrary?
Respectfully,
– Michael
Michael,
I’m willing to accept both methods are capable of leveraging returns. I wouldn’t describe either approach as safe. In addition it can be perfectly reasonable for someone to own an aggressive unlevered portfolio. In some scenarios the more aggressive unlevered portfolio might even be optimal.
I’m just arguing for a disclosure that the aggressive portfolio may have twice the volality but does not guarantee twice the return, even if that was what was originally targeted. There is ample evidence to suggest the more aggressive portfolio does not always live up to expectations.
Anthony,
Indeed, I don’t mean to suggest my particular portfolio examples here are the ultimate final recommendation about how to create a volatile portfolio.
But I do think it’s a failing for us that we’d rather have the “easy” discussion about how to use leverage, instead of the “harder” discussion about how to effectively design more aggressive portfolios. And I think a lot of clients (and planners?) are fooling themselves into thinking their strategy is not aggressive, just because the PORTFOLIO part is conservative, even though there’s a ton of leverage involved.
And viewed from another perspective – if we could work on creating a portfolio that accomplishes the client’s investment goals WITHOUT leverage, why is that not an appealing client conservation? And if we as planners need to learn more about portfolio construction to learn to do that, then let’s learn, not naysay and use leverage instead!
Respectfully,
– Michael
Anthony,
As a sidenote, I also still disagree with the baseline here.
Why do we have to disclose that a volatile portfolio might not generate the same returns as a leveraged one? Why don’t we have to disclose that a LEVERAGED portfolio with a mortgage might not generate the same returns as a “merely” more aggressive unleveraged portfolio?
I don’t disagree with the importance of disclosing and discussing volatility and risk. I just don’t understand why we insist that using leverage is the baseline against which all other strategies are compared?!
Respectfully,
– Michael
Mark,
The point here is not about good debt vs bad debt (although in reality, I don’t think of good debt vs bad debt – I think it’s about good INVESTMENTS vs bad investments!). It’s that using debt to purchase has RISK, and specifically that we seem to ignore the risk of debt when we already have it (the existing mortgage) but not when we’re adding it anew (taking out a fresh mortgage to buy stocks).
Real estate investors use leverage because they find the risk appealing for the potential payoff it provides. On the other hand, real estate investors also have some of the most SPECTACULAR bankruptcy implosions if the debt gets out of hand. That’s not to judge it as good or bad; that’s simply to acknowledge that debt and leverage have risk and tradeoffs.
But the point here is that unlike the real estate investor, the stock investor can produce comparable levels of return WITHOUT the debt, just by buying more volatile investments. Which raises the question about why you’d use debt – and the attendant risks – if you didn’t have to in order to get the same return!
But yes, clearly if you aren’t confident in your investments and don’t think they can outearn your borrowing cost in the first place, it’s clearly a losing proposition. The point here is the distinction between buying investments that will slightly outearn your borrowing cost with a lot of leverage, versus just buying more volatile high-risk/high-return investments in the first place and not bothering with the leverage at all!
Respectfully,
– Michael
Michael,
With all due respect, Michael (hereafter referred to as WADRM), as I see it, you have this backwards – if I own and owe $500k on my home, and hold a $1M portfolio – then your 2x strategy on a $500k portfolio with no debt (ignoring the very real world issues associated with implementing the 2x strategy pointed out by Anthony above) replicates the overall risk profile of the $500k/$1M structure – but ONLY for the moment. The 2x strategy would keep the overall risk at 2x, while the “keep the mortgage” strategy offers a decrease to the overall risk profile as the portfolio and/or real estate values increase (at a portfolio value of $2M, I have a less levered overall structure). The reason that after adjusting for a year’s debt service you have more money in your example above, is simply because you are holding more, not less total profile risk. In fact, at any portfolio value above $1M, you have less than 2x total profile risk.
It is also true that any portfolio value below $1M (in your example) produces a total risk profile GREATER than holding the no-debt, 2x portfolio. Ironically, this is where I have the most trouble with your analysis… and where I look forward to much greater discussion.
Owning and owing $500,000 at the point where my portfolio is worth only $500,000 means I will require a 4x portfolio to match the (mortgage) leveraged position. At a $250k portfolio, 8x would be required. Given the difficulty with tracking error and volatility drag associated with even a 2x strategy, I’m hard pressed to imagine how I might keep a client in such a portfolio (“don’t worry, ignore all that talk about tracking error, in the end THIS portfolio is designed to return 8 times that of your previous portfolio”).
You may rightly say “this is my point, we are taking on far more risk than we even realize through our casual, everyday use of mortgage debt,” but WADRM consider this:
In the framework you’ve offered (contribute to 401(k) or accelerate mortgage payoffs), you have appropriately focused the discussion of downside risk around the possible loss of income and inability to make one’s mortgage payment. But for the period where one is accelerating the amortization of their mortgage debt, there is actually MORE risk in having the shorter amortized payment. That is to say that in the 1st ten years of (say) a 10-year mortgage there is more risk associated w/job loss than there would be if the required payment were amortized over 30 years. When discussing the risk of job/income loss, our attention should include the increased risk of larger required mortgage payments in the early years, not just the increased risk of (still) having a mortgage payment in the later years. In an apples-to-apples comparison, it should also include the additional emergency fund savings required to cover the additional monthly payment.
Lastly, you original post on this subject offered that business debt be considered differently than personal (mortgage) debt – owing to the fact that businesses could in effect default with fewer consequences. Again, in your framing this conversation in terms of accelerate mortgage payoffs Vs. contributing to 401(k) it seems worth mentioning that money in one’s 401(k) is protected from creditors… and that original acquisition indebtedness is generally not subject to recourse beyond the (then) value of the home. Morality aside, the person who lost their job after paying off $200k of their $500k loan only to see their home decline in value to $300k, may rightfully wish they were in the position to walk with that money inside their 401(k).
WADRM,
~ your friend, Don
Don,
A few quick thoughts here…
– Yes, the inherent leverage of the strategy will shift over time as portfolio values change. But I don’t see that as a problem. That’s easily addressed. We rebalance the portfolio (and/or otherwise adjust to keep volatility in line). That is already a standard strategy for portfolio management, and it would be equally relevant here. It’s just that you might rebalance based on volatility of the WHOLE balance sheet, not volatility of “just” the portfolio (which, arguably, we should be doing already).
– Yes, again, the inherent leverage shifts over time, but frankly the mortgage strategy seems LESS favorable in this instance. Why would you want a strategy that implicitly REDUCES leverage when things are going WELL, and INCREASES leverage when things are going badly? In point of fact, that’s the exact trap that a lot of banks ran into, with serious problems. MF Global has already acknowledged that part of the reason they were misappropriating money from client accounts is because they were trying to dig out of a hole of exponentially increasing leverage as their capital base eroded. Getting caught in a declining-capital-leverage-spiral is an ugly thing. Why would we want to subject clients to that? At least with a volatile-but-unleveraged portfolio, we can manage the risk on an ongoing basis in a way that doesn’t cause risk to skyrocket just AS things are going badly!
– On the third point, regarding the huge volatility necessary to replicate more leveraged strategies… yes “this is my point, we are taking on far more risk than we even realize through our casual, everyday use of mortgage debt.” ’nuff said. 🙂
– Although I would advocate accelerated payoff strategies, that does NOT necessarily mean I would advocate shorter loan commitments. COMMIT to long periods with low required payments, and then voluntarily accelerate the payments. You will pay a slightly higher interest rate, but that’s the cost of having the FLEXIBILITY to keep the leverage, if you want to, or if you NEED to because you need to redirect the cash flow elsewhere. So in my context, I *hope* to pay off my mortgage in 10-15 years at the most, but I *have* a 30-year mortgage so I can drop the payments dramatically if I want/need to. If you pay off over an accelerated time period anyway, the difference in interest paid between a 15- interest rate and 30-year interest rate is not very dramatic, anyway.
– So in essence, it’s minimize payment commitment, maximize payments when cash flow allows, drop back on payments when necessary, keep a home equity line of credit so you can re-borrow if you REALLY need to, and match the deleveraging of your balance sheet with increasing volatility in your investments if you really want to take that much risk.
Respectfully,
– Michael
Michael,
On the third point… I have conceded your point (“we are taking on far more risk than we even realize through our casual, everyday use of mortgage debt.”), but you have not yet fully addressed mine (that the volatility/portfolio risk necessary to replicate the mortgage-levered strategy in the early years renders the latter the only viable strategy available to those who might someday be in the position you describe: having the ability to pay off their existing $500k loan using their $1M portfolio).
Accounting for a 25% interest deduction, someone choosing to pay off their 5%, $500k mortgage in 15 years must earn 12.65% to match the portfolio value of someone amortizing the same debt over 30 years *earning just 8%* – who instead of waiting until year 16 to begin investing their *entire* mortgage payment, starts investing at the beginning of year one, using only the difference in mortgage payments (15 vs 30) plus the additional tax savings.
*12.65% over 15 years vs 8% over 30.* Again I will concede that *over the entire* 30-year period, the risk profile(s) of these alternatives is (almost) identical. “Almost” owing to the fact that there is some portion of this arbitrage attributable to the interest tax deduction (in this case 114 bps, or the difference between 12.65% and the 11.51% that would be necessary to match the 30-year amortized, 8% return, invest-the-difference scenario if the interest were not deductible). Unless CAPM accounts for this “tax arbitrage,” the 114 bps amounts to an additional fee levied against the 15-year amortization/salt-it-away later alternative.
But while I will concede the two strategies share similar risk characteristics *_over the entire time period_,* the point I am driving at is that the leverage necessary to achieve these results is *front-loaded,* that is to say the risk is focused more heavily on the early years of the strategy (when the client has *less to lose and more capacity to recover*) in exchange for a total risk profile that is lower in the future (when the client has less capacity to recover and a larger portfolio to protect).
Relative to the 2x strategy this is like saying “I’m willing to take 4x, 8x, 10x total risk in the early years – while I’m building my wealth – so that I can take 1/2x, 1/4x in the later years, once I have already acquired my wealth.”
Getting caught in a declining-capital-leverage-spiral is indeed an ugly thing. But the way one gets himself into such an environment is to leverage-up during good, but unsustainable times. Neither of us (no any of the commenter’s so far) are suggesting anyone leverage-up. The argument is about how quickly (or not) to leverage down – and specific to my point, when (if ever) in the lifecycle is it appropriate to hold more leverage, and when is it appropriate to hold less?
And the fact remains, the early years are subject to the same job loss risk whether or not one employs an accelerated payoff strategy. It’s only at the where the (accelerated) mortgage is paid off (and the non-accelerated is not) that there is less risk of job loss for the accelerant – the period say between 10/15 years and 30.
For your consideration… 🙂
Don
Don,
The strategy only requires increased leverage in the early years because that is what the mortgage created in the first place!
The investing-on-mortgage strategy is using a huge amount of leverage in the early years, and dialing it down later by slowly amortizing the mortgage.
If you don’t like the fact that the alternative requires such huge front-loaded volatility to make it work, then why are we accepting of the investing-on-mortgage that is creating the front-loaded leverage in the first place?
In other words, all the leverage characteristics you’re drawing as negatives are not a result of the volatility-driven portfolio I’m suggesting. They’re a result of the investing-on-mortgage strategy that the profession is using as a baseline. If we don’t like it, why are we taking on huge leverage in the first place?
If you simply prefer the 10X/8X/4X/2X/1.5X deleveraging that is implicit in buying stocks on mortgage, replicate it with volatility in the portfolio. If you don’t like it, then apparently you don’t like your mortgage, either?
Respectfully,
– Michael
Michael,
My point is that you *can’t* replicate the 10x/8x/4x/2x/1.5x deleveraging — not at least at the beginning (10x/8x/4x) of that lifecycle — any better than you can with the mortgage (owing to the deductibility of the interest). Your own framework proves this out (no difference in total profile risk between the *theoretical* 10x/8x/4x portfolio and the corresponding mortgage enhanced 1x portfolio). Please do not suggest you can get there with long-term options – where the drag on S$P 500 returns has been demonstrated to average more than 500 bps, not some “expense ratio-like” trivial cost.
Even if you *replicate* the exact leverage of the 1x plus leverage strategy, you’d still need to convince the client to accept the new strategy and associated brain damage. Not to mention the (now) more obvious volatility. Volatility that Dalbar, Morningstar, Fidelity and others point to as the most meaningful expense investors will likely encounter.
I appreciate the framework (theoretical or not) that brings forward the total risk/return profile in a more holistic manner than most would otherwise think about, I’m just not enamored with your conclusion.
WADRM,
Don
Don,
If our greatest problem as financial planners is that we’re having trouble coming up with OTHER ways to help clients invest with 10:1 leverage, then maybe we need to re-evaluate how much we’re telling people to save and invest in the first place?
I don’t mean to belittle the conversation – after all, I’m the one who started it – but it does seem somewhat absurdist to me that we’re debating the best way to help clients invest with 10:1 leverage to achieve their goals.
If 10:1 leverage is really what’s required to succeed with the plan we’re creating for them, maybe we need a new plan… just on principle? 🙂
Respectfully,
– Michael
Don,
It’s also worth noting that while the leverage is the greatest in the early years, it’s when the account balance is smallest (as you point out to some extent).
What that means, though, is that while there’s a lot of leverage in the early years that’s hard to replicate with volatile investments, it’s also a lot of leverage on very little actual money, which means it also doesn’t actually have a huge impact.
In other words, I could try to figure out how to replicate 10:1 leverage on a few thousand dollars in my account… or I could just try to save another thousand dollars, which may have more impact anyway.
By the time the accounts are large enough that the leverage has a more significant dollar impact, the magnitude of the leverage is no longer as significant.
Food for thought.
– Michael
Not all debts are bad. There is good debt and bad debt. If you can make a higher return on your money than the interest you pay on your mortgage, why would you want to pay off your debt instead of investing the money?
Let me give an example. Every real estate investor leverage debt to invest in property. It is just stupid not to do so, which is why no professional real estate investor would pay off a piece of property in full.
Is it any less risky to invest in real estate than it is to invest in stocks? Both can be just as risky if you don’t know what you’re doing. So if it’s smart to leverage to invest in real estate, than why is it not smart to leverage debt to invest in stocks or any other asset class?
The question is, do you know what are doing? Are you confident of your own investments? Although there’s no 100% guarantee, but if you’re not at least confident in your own investments, why invest?
Mr. Kitces,
I want to thank you for not only having a post that 100% directly addresses an issue I’m personally wrestling with (I’m one of the guys on the other end of the phone with the folks that have been writing here 🙂 ), but for creating a post that has created an internet record for the highest quantity of valuable discussion points with no spam or trolling I’ve ever seen! 🙂
Our issue is that we have 20 years left on a 30yr. fixed with a balance of like $117K on an initial purchase price of $417K (current value is most likely $390-$400K).
We have ~$400K in “untouchable” 401K’s, trusts, etc. and about $200K in regular stock and bond funds that are pretty low beta/risk. About $60K in cash/money market as well.
I love your mental model of looking at this as a balance sheet question and not just an investments vs. debt question.
My wife and I are both in our early 40’s. I really want to pay off our mortgage because I clearly see the writing on the wall with regards to the global financial picture. There’s just too much bad debt globally. There’s significant risk that in the next two years I may lose my job and our investments tank.
My goal, thus leading to my desire to pay off, is based on a point of elimination of risk (not just financial risk, but “life” risk… Not having a house or place to live or a means to pay for it if the SHTF in a big way). (BTW, I’m a huge disciple of Taleb’s Black Swan ideas and they really helped us in 2007-2008. I’m much more comfortable risking losing out on an upswing than blowing up on a downswing.)
Our PM is really against this with the old arguments of interest on mortgage (~5%) is lower that “forecasted” returns of even our moderate portfolio we would forgo by paying off the mortgage, tax write off’s, etc.
I understand the argument from a purely academic, objectivist standpoint, but I still think, looking at the current world financial situation, the most prudent thing to do is to totally deleverage. Get out of the mortgage, and if I’m wrong about the global financial situation and things start rocking in 2-3 years, I’ll just ratchet up the risk level in the remaining investments.
Anyway… That’s what this laymen pleb got from taking my starting point of wanting to pay off my mortgage and reading this amazing blog.
If I’ve fallen off the rails with my thinking, I’d love someone wiser to point it out before I pull the trigger! 🙂
Freezing cold take. My sympathies, given the power of hindsight.
Michael,
Thanks for the very interesting blog entry. I agree with you that debt is often not the best way to boost returns, but increasing risk level of the overall asset allocation on the efficient frontier is often more effective in boosting returns.
Having said that, I like debts for the liquidity they provide. For example, a margin loan at 0% for 3 days (between the trade day and the settlement day) isn’t a bad deal. Neither is credit card debt at 0% for 30 days. The same can be true for a mortgage at a fixed low rate for 30 years.
I think you missed this one Michael. Mortgage debt is far superior to margin debt. It is non-callable, long-term, low rate, and tax-deductible. The solution you’re advocating instead is to increase a portfolio’s beta from 0.5 to 1.0. Aside from the fact that someone most interested in portfolio growth will likely have a portfolio with much higher beta than 0.5 to start with, your analysis ignores the behavioral issue. In my opinion an investor is far less likely to bail out at market lows with a portfolio decline of 25% versus 50% and is unlikely to take the mortgage into consideration in the decision of whether to stay the course or not. How it feels matters, and ignoring that is folly.
Not to mention the tax arbitrage of investing in a tax-advantaged account instead of paying down a mortgage. I think you can make a decent case for paying down a mortgage rather than investing in a taxable account, but passing up a 401K or Roth IRA contribution for it? That’s a lot tougher.
White Coat,
The leverage factor here can be anything you want; the 0.5 to 1.0 beta was just an example that makes the math a little easier to follow.
The point here is that if the investor is going to panic because of a 50% portfolio decline, they should be just as panicked by a 25% leveraged portfolio decline, because the percentage of NET WORTH being lost is the same in both scenarios (except the 25% leveraged decline is actually worse, because it’s also higher cost due to the interest being charged on the loan!).
Yes, I will certainly agree that IF someone wants to massively leverage their portfolio, mortgage debt is superior to margin debt, for all the reasons you note (though margin investment interest can be deducted, albeit with restrictions, so that’s sometimes a wash).
But comparing the mechanics of margin vs mortgage debt is just about “the best way to finance a massively leveraged portfolio” – it doesn’t change the fundamental point that the same risk-return parameters, RELATIVE TO NET WORTH, can be accomplished WITHOUT the leverage and borrowing costs in the first place, as illustrated here.
Notably, many of the most catastrophic financial disasters in history were specifically born of this approach – with a “well diversified” portfolio that was deemed to be safe, leveraged to the hilt, and only after something bad happened was there acknowledgement that while the portfolio may have seemed diversified and “safe” that with leverage, it doesn’t take being wrong by much to wipe out investor equity altogether (whether it’s Long Term Capital Management or a portfolio of CDOs).
And if you’re not comfortable with the portfolio WITHOUT the leverage, pretending that it’s more comfortable WITH the leverage – when in fact the latter is both riskier AND more expensive – should not be a recommended solution just because it conveniently fits our mental accounting biases! 🙂
– Michael
I agree it is mental accounting. I also agree that being highly leveraged in any situation is probably a bad idea (such as the person with a $600K mortgage and a $400K portfolio.) But ignoring behavior is folly, and mental accounting affects behavior. I don’t think people make the decision to stay the course or not based on net worth. They do it by how they feel watching the numbers in their investment account drop statement after statement. They don’t think about the mortgage.
Another point, perhaps best illustrated by an example. You’ve got a 35 year old guy with $10K and the opportunity to put that $10K into his 401K or use it to pay down his mortgage. Let’s say he has a $300K portfolio and a $200K mortgage, so not ridiculously leveraged. His marginal rate is 33%. He expects to withdraw that money at an effective rate of 15%. The 401K costs are low and he expects long term returns on his 80/20 portfolio of 8%. The mortgage is 3% (2% after tax). Are you really suggesting he move the portfolio to 100% equity and maybe add a little small value stock or emerging market stocks and then send his $10K to his mortgage holder instead of putting the $10K into the 401K? Seems kind of dumb to me. You’re missing the tax arbitrage opportunity, you’re missing the increased tax-protected space, and you’re missing out on the leverage benefit of borrowing at 2% and earning 8%. Surely bumping up your return by 0.5% or so with the riskier portfolio isn’t going to make up for missing all that.
The repayment of the 2% borrowed is risk-free. The 8% market return is not. Wish I had completely paid off my mortgage before the 2007 crash. I borrowed a lot of money for the privilege of losing much of it in the stock market crash.
Great post Michael. The last several years I have discussed with clients paying off (some or all, reminding them its not an all or nothing scenario) mortgages as a “synthetic” bond position, and increasing their risk in their portfolios as a result. Mental asset allocation, with pitfalls of liquidity for a guaranteed savings on the loan interest.
As I approached 50, and I saw we were closing in fast on our “number,” the balance sheet wasn’t lost on me. It was skewed toward pretax accounts, but still had years to go on the mortgage. I follow your math, and on an intellectual level, agree with your conclusions.
In our case however, the $250K left on the mortgage is $375K in a 401(k) account. The combination of a child, high property tax, an HSA (I am working part time, doing a job I love, thus the insurance and HSA), and years of carried ATM credits, my tax bill will be nearly zero for the next 10 years.
If the saving were post tax, I’d look at this as you did, and using the cash portion of our allocation, kill the mortgage, as even the 3.5% mortgage savings beats what I expect cash to pay for the next 5 years. But in my situation, the mortgage and the payments aren’t a burden, but the cost to pull the money our of our retirement accounts to pay it off isn’t worth it.
I am very intrigued by the article but left a little confused-could you help me clarify my understanding?
The three cases seem to have two variables moving rather than one. It seems A and B are not investing their entire portfolio in the stock market. I would think in your years in your example when the market was up 10%, C makes $50k and A and B make $100k (minus $25k of interest payments so net $75k). To me this illustrates that taking out a mortgage and arbitraging is a great strategy and you’ve convinced me in this article that I should consider margin loans as well? I seem to be missing some of the nuance. Thank you for youe help.
It’s very difficult for an advisor who charges a client a percentage of assets managed to advise that same client to withdraw funds from the managed account to payoff mortgage debt.