December 9, 2020
Right at the start, let me point out that, no, I’ve not gone to the bad side! I will not try to sell any actively-managed funds here. If you’re a part of the passive investing crowd, which is a large portion of the FIRE community, you might find the title a bit “click-baity.” Because the thought process of the average passive investor would go like this:
- Underperforming the VTSAX is a non-starter. That’s highly undesirable. The only assets we’d ever consider are those with an expected return equal to or larger than the VTSAX!
- But the problem is that due to efficient markets, nobody can beat the market!
- If we intersect the two sets above, i.e., constrain ourselves to what’s both desirable and feasible we’re left with the VTSAX (or whatever close substitute you might pick, e.g., FSKAX from Fidelity).
That line of reasoning has some advantages: it has probably convinced a lot of folks to get rid of their irrational fear of the stock market and many have benefited from low-cost index investing instead of wasting money on actively-managed funds. My concern here is that I think that this thought process of “nobody can beat the market” is overly simplistic and (literally) one-dimensional. Of course, there are ways to beat the market! Here are eight ideas I can think of…
The easiest way to “beat the market” is so trivial, I am almost embarrassed pointing it out. Since the stock market goes up over time (at least on average) you can easily beat the market through leverage! Let’s see how this would work in practice. First, we have to graduate from the naïve, one-dimensional view. Let’s add another dimension: risk and look at the risk vs. return tradeoff. If you have access to cheap leverage, say by trading futures then you can easily walk up that expected return. If you agree that equity index futures have returns equal to the total return minus a risk-free rate then we can extend the risk-return tradeoff along the straight line connecting the risk-free asset and the 100% equity portfolio.
As always, there are several caveats:
- The extra risk may not be worth the extra return: Let’s face it, for most folks a 100% equity portfolio is already too much. But for very young investors who just start on their path to investing, one could make the case that investing in equities on leverage makes sense. Two researchers from Yale make this point, and it all has to do with Sequence Risk, i.e., over the average investor’s lifecycle, you’ll have way too little equity risk early on but likely too much equity risk right around retirement. The solution? Leverage when you’re young, and then a stock/bond glidepath toward retirement.
- Legal, regulatory and personal constraints: Forget about being comfortable. There are some regulatory restrictions on using leverage, for example in retirement accounts.
- Costs: This would be the least of my concerns because you can get exposure to leveraged equity beta very, very cheaply through equity futures. But for small accounts, if you have to rely on leveraged ETFs, the additional cost through the expense ratio of such funds would make their use a lot less appetizing.
- Skewness risk: This is a big concern for leveraged accounts. If the market goes down by 10%, you’d need a recovery of 1/0.9-1=0.111=11.1%. If you had invested with 3x leverage and your portfolio is down by 30%, then you need a 1/0.7-1=0.426=42.6% return just to get back to zero. That’s more than 3x the unleveraged value of 11.1%! If you had invested with 10x leverage you’d be wiped out and you’ll never recover. So, leverage has to be used very cautiously!
By the way, this leverage issue comes up all the time. Especially now that mortgage rates are so low, a lot of readers have been asking me if it’s wise to keep the mortgage in retirement. I wrote a post on this a while ago (SWR Series Part 21) and the message is still the same. Mortgage rates are certainly lower now than when I wrote the post. But so are equity expected returns, thanks to a very lofty CAPE ratio of well above 30. It’s probably a wash, so we are still sitting here very comfortably here with our mortgage-free home. But again, if you’re still working and especially if you just started to invest you should certainly consider not just 100% equities but even equities plus leverage.
2: Better risk-adjusted returns through diversification
The previous point, beating the index through leverage, demonstrates how incomplete the fixation on average/expected returns is. Of course, you can beat the market – on average and in the long-run – if you lever up the whole thing. But you’ll also expose yourself to a lot more risk. That brings me to the next point: you can easily beat the market if you target a different objective function that takes into account not just the expected return but also risk. Let’s look at the following numerical example:
- Assume 6.0%, 0.9%, and 0.05% expected returns for stocks, bonds (10-year Treasury bonds), and the risk-free asset (e.g. money market account, idle cash balance interest, 3-month (or shorter CDs, etc.). A bit of a conservative equity expected return and the bond yields roughly in line with the current yields.
- Assume 15% and 6% annualized risk (standard deviation) for stocks and bonds. And 0% for the risk-free asset (by definition!)
- Assume stocks and bonds have a correlation of -0.3, which is close to the actual correlation in recent history.
If we look at the return stats of stock/bond portfolios, where stocks are x% and bonds 100%-x%, then you achieve…
- The maximum expected return with 100% in the highest expected return asset. Obviously!
- The maximum Sharpe Ratio, calculated as the difference between the portfolio and return and the risk-free asset return divided by the portfolio risk (=risk-adjusted return), at around 40% equities, 60% bonds.
- The minimum volatility on the efficient frontier is at around 20% stocks. (but admittedly, that’s also pretty close to minimum expected returns!)
So, depending on your exact objective function we should most definitely consider allocations with less than the equity return. Maybe not so much when still accumulating assets but certainly in retirement, which brings me to the next point…
3: A 20-40% bond allocation lowers the “tail risk” in safe withdrawal rate simulations
People often ask me whether it’s a good idea to stay at 100% equities throughout retirement. The only people who should do that would be the early retirement bloggers with enough cash flow from their blog and other business ventures or their spouse’s income so that they don’t have to worry about Sequence Risk. The rest of us will be better off with at least some additional diversifying assets.
I quickly simulated a few different asset allocations, between 50% and 100% stocks, the rest bonds, over a 30-year retirement. See Part 28 of the SWR Series for the link to the Google Sheet. Despite the lower expected return, you’ll do better with a moderate allocation to bonds. In other words, in the final 20-40% of the portfolio, you do better with bonds from a Sequence Risk perspective. The success rate of the 4% Rule is highest and the failure probability is the lowest at around 70% stocks, 30% bonds.
4: Tangency point plus leverage
Remember #1 (leverage) and #2 (better risk-adjusted return) above? We can now combine the two and reach an expected return above equity returns while keeping risk at the same level or lower than with 100% equities. I wrote a blog post on this topic a long time ago (“Lower risk through leverage“) and this is the perfect opportunity to revive this vintage article.
Let’s go back to our numerical example again with the expected for stocks/bond/cash at 6%,0.9%,0.05%, and the same risk parameters. Not too far from today’s environment. If we plot the efficient frontier and find the maximum Sharpe Ratio as the tangency point of the straight line starting in the “Cash” point, then we could keep investing in that max-Sharpe portfolio through leverage and get into the area to the left and above the original efficient frontier (i.e., higher return and lower risk). We can even mark that green triangle up and to the left of the 100% equity allocation. Along the hypotenuse of that triangle, we observe higher expected returns and lower risk than the stock portfolio! Sweet!
And you know what’s the beauty of this approach? It’s still purely based on the index returns. No stock picking!
Also, if you wonder if this is just some theoretical pipe dream that would never work in practice: When I worked in the industry, we indeed offered a product (to institutional investors only!) that’s based on this approach. It had a 30-year return history of beating the S&P 500 by 3% a year. Admittedly with a ton of additional bells and whistles, but the core of this strategy was based on the idea of an unconstrained portfolio optimization problem that allowed for some modest leverage to get you past the constrained efficient frontier. So, this idea really works in practice! Not every year, but on average and over the long-term.
Notice that in items 1-4 above, I still operated in the confines of the broad index only. No stock picking! Let’s go beyond that and look at what can be done by deviating from the broad index weights. We will not go all the way to picking individual stocks but maybe go “half-way” by picking certain styles within the broad market. This brings me to items 5 through 7…
5: High-beta stocks
OK, so let’s assume that leverage (see #1 or #4 above) isn’t your thing; sounds way too dangerous! Here’s a way to achieve leverage through the “backdoor” without ever explicitly employing leverage. Simply buy stocks with built-in leverage in the form of higher-than-average market beta. With beta, I mean the (regression) slope estimate with respect to the overall market. Higher beta stocks have higher expected returns. And no, this a no violation of any efficient market hypothesis. Quite the contrary, it is a direct implication of the CAPM model according to which the expected return E(R) of an individual stock should be related to the risk-free rate RF and the market expected return E(Market) via:
E(R) = RF+ beta*[E(Market) – RF]
Thus, if beta>1 then E(R)>E(Market) and you “beat” the market, at least on average.
How do you find the beta of stocks? Yahoo Finance displays the beta right on the stock summary page:
In Google Sheets you can use the function googlefinance:
… gives you a value of 1.20 (as of 12/5/2020).
Fidelity also has a great stock screener to look up the beta of a whole list of stock tickers. So, simply screen your watchlist for stocks with betas greater than 1 and you get leverage and the boost in the expected returns without ever having to mess around with margin loans or futures trading. Sweet!
And a caveat: The value of beta depends on the horizon over which you run that CAPM factor model regression. I’d normally look at the 5-year beta. Shorter horizons create very noisy estimates! I am not sure what methodology and what horizon Google is using. Yahoo Finance evidently uses a 5-year horizon with monthly return data.
6: A market anomaly: “Betting against Beta”
After referencing the plain old CAPM model, people will be quick to point out that there are all sorts of well-documented holes in that CAPM theoretical construct. Glad you mention this because a lot of the CAPM abnormalities can then be used to – you guessed it – beat the stock market. One deviation from the CAPM that I found really intriguing is this: Researchers Andrea Frazzinia and Lasse Pedersen pointed out that the capital market line in actual data tends to be flatter than predicted by a simple one-factor-CAPM model. Their paper is called “Betting against Beta“. In other words, high-beta stocks have expected returns a little bit below the blue capital market line, see the chart below, while low-beta stocks have expected returns a bit above the capital market line.
The researchers document this phenomenon empirically, not just in the U.S. but in essentially all other country indexes and even other asset classes. Also, I should stress that low-beta assets do not necessarily do better than high-beta assets. The line between the purple dots is still sloping upward. But low-beta assets do better than what the CAPM would predict. An overweight to low-beta stocks and underweight to (or even shorting) high-beta stocks would give you excess returns or at least excess risk-adjusted returns.
The researchers also provided a pretty compelling theory for this CAPM abnormality. Too many leverage-constrained investors try to use the high-beta play to get higher returns without the explicit use of leverage (see item #6 above). That drives down the expected return of high-beta stocks and leaves more for the investors willing to invest in the undesirable low-beta stocks.
7: Other market anomalies: “Size” and “Value”
Well, as I hinted at in the previous point, in that CAPM equation, the market factor may not be the only explanatory variable of stock returns. Long before the BAB factor was discovered, people showed that in addition to market exposure, the size of the corporation and the price to book ratio account for a lot of the equity return volatility. And so the three-factor model was born (market beta, size, and value). And quite intriguingly, the small stocks and value stocks have outperformed the overall index since 1926, which is as far back as the Fama-French factor database goes.
Personally, I am skeptical about a persistent and continued outperformance of the small stock and value stock and small-cap-value style. See my 2018 post My thoughts on Small-Cap and Value Stocks. But you’ll find a ton of people online who will swear by these two factors. In fact, the two-factor premia are so engrained in the thinking of finance professionals, the idea of a small-cap and value premium has become almost a “natural law” in finance. For example, if you look at the JP Morgan Long-Term Capital Market Assumptions, there’s always a higher expected return for small-cap and value stocks, see the table on page 118 of that report. But keep in mind that the higher expected return also comes with a higher volatility estimate!
8: Adding other uncorrelated assets
One important issue that often gets lost in the VTSAX-mania is that some assets, when compared side-by-side to the passive index fund, might seem unattractive. Homeownership, for example, gets a bad rep in some circles in the personal finance community, but that’s mostly because people don’t know how to calculate returns correctly. Homeownership can be significantly more profitable than the stock market (my own experience here: “My best investment ever: Homeownership?!“). But even if my homeownership return had lagged the stock market by a percentage point per year, it’s still a decent low-risk investment. Especially in retirement, you will benefit from the stable “income” in your own home in the form of not having to pay rent. It’s one of the best Sequence Risk hedges, even if the homeownership return may lag the equity return by a little bit.
The same is true with my other non-equity-index strategy: Selling put options, see my 2019 post on what exactly I’m doing. Over the long-run, I’ve slightly outperformed the S&P 500 with my strategy. But in 2020 I’m lagging by 2 percentage points. Is that a failure? No, because I also had significantly less risk: the S&P500 volatilely was about 25%, my put strategy only 11%.
I hope the post wasn’t too much “false advertising”. If I had to put the eight “beat the market” ideas into different buckets I’d use these:
- Bucket 1: The “beat the market” suggestions #2, #3 still have a lower expected return, but at least they offer better risk-adjusted returns than the stock market.
- Bucket 2: #1, #5, and #7 have potentially higher returns but that “beat the stock market” comes at the cost of higher risk.
- Bucket 3: Only #4 truly offers a higher return and lower risk, at least in expected terms. #6 and #8 might get you there too. Though, to be conservative, I’d still put those in the first bucket “lower expected return but better risk-adjusted return”.
If you’re skeptical, I don’t blame you. Stick with your VTSAX. But for the rest of us who are interested, there are indeed a few things we can do beta, uhm I mean, beat the market!
So much for today! I will likely take some time off from blogging for the rest of the year. So, have a great Christmas season everyone, and a good start to the New Year!
108 thoughts on “How to Beat the Stock Market”
I’m going to be linking to this post a lot. 🙂 Thanks!
Glad you enjoyed it! 🙂
Great blog, as ever, Karsten! Thanks for your fantastic insight in a rational, mathematical approach of beating the market. Just a question: was a (beta) pun intented in this sentence at the bottom of your blog? ‘But for the rest of us who are interested, there are indeed a few things we can do beta the market!
Hah, that was a typo, but I changed it into a pun now! 🙂
Great post! Would love to know your thoughts on implementing the backward planning approach to asset allocation based on size of assets…any examples/insights on how to implement this?
Big ERN: amazing post as usual! I have to read it 3-4 times just to fully understand it all! Thanks for sharing your expertise and knowledge with us.
You mention leveraged ETFs briefly in your post. It would seem to me that the use of 3x leveraged index ETFs (UPRO, TQQQ) may have role in trying to outperform the market as well, particularly if there is a fairly obvious market direction ahead (i,e., after a correction, when tax loss harvesting, etc.) Yes, the ER is pricey (about 0.9 for most of them) but there is some potential reward for that cost and for those of us a bit intimidated by other forms of leverage, it seems a bit more palatable as you can only lose what you put in and the reality is that the SP500 and NASDAQ will (always?) recover eventually. I read a lot about time decay with these ETFs, but this seems to really be mostly an issue when there is persistent up and down volatility and not movement primarily in one direction.
Just curious on your thoughts on the potential role of these ETFs.
Happy holidays to you and your family and thanks again for what you do!
The main issue I have with the leveraged funds is the cost. However this can be overcome once one has even a little bit of capital. A quite similar portfolio can be created using micro e-mini futures at a fraction of the cost. The notional size of one e-mini S&P500 future sits around $18k right now, so to hold a 3x leverage position you’d about $6k. The minimum margin on this contract is around $2k so you’re relatively safely above that. Instead of paying around $55 per year ($6k of UPRO) you can buy one micro e-mini and roll quarterly for like $6/year. The best part is with a 10x larger portfolio you can use the regular e-mini and the cost stays at $6/year!
Yes, the ETF approach is only for really small accounts. Should work much better in larger accounts using futures and options.
For small portfolios, the leveraged ETF probably work well. Once you get to six-figures and certainly 7-figures, you’ll be better-served with using futures/options.
Was starting to worry about you, so glad to see post. Great stuff as always.
Haha, thanks. 🙂
Huzzah. Someone needs to create a High Beta ETF for this reason. Do it, ERN! I’d invest 😉
Or maybe do 2 ETFs, one with high-beta and one betting against beta. One of them will do really well! 🙂
THX for your great article (as always).
POTION No. 9 😉
How about a suggestion #9 to beat the market:
– leverage the stock market only if there is a high(er) chance of mean reversion
– wait until the market (broad index ETF) has lost 25/35/45% from last top
– go in tiers into the market with moderate leverage (1,1/1,2/1,3x)
– say to your self: this leverage is the total of my preferred saving plan rates for the next 6-24 months
– according to your risk level you can pay back the credit in the next 6-24 months before applying real new (earned) cash into the stock market
In our case, it worked somehow well (run out of money way to early 2008/Mrc09 but then after not).
What are your thoughts?
All good and valid points. However, you’re now getting into market timing and you would have lost a lot of the upside potential during the long bull market.
But I like the idea of calibrating the leverage size to the contributions over the next x months.
I’ve been digging into this more recently, especially the idea of leveraging up bonds to get risk parity (pretty much approach #4). If people are looking for funds that do this, the bogleheads forum has covered this extensively and has a few favorites. For taxable accounts, NTSX does a 90% stocks 60% bonds leverage strategy (i.e. 50% leverage on a 60/40 portfolio) that uses a spread of treasury futures to get the additional bond exposure. It’s relatively cheap (0.20% ER) and tax efficient. For my IRA’s, I’m considering PSLDX which uses a mix of strategies to get 100% stocks and 100% bonds exposure. Their duration risk is much higher though, so not for the faint of heart in an interest rate increasing environment. But their performance over the last few years has been bonkers. Very high yield though so not suitable for taxable.
NTSX looks attractive. Very nice performance this year. To get closer to the max-Sharpe, I’d prefer the 80/120 allocation over the 90/60 but it’s definitely a good start.
Thanks for the link to PSLDX. Hadn’t heard of it before. Good strategy to replicate and save the 1% fee. 🙂
Haha yea, I guess rolling your own is cheaper. I like to keep it simple so don’t mind the 1% ER (technically its a 0.6% ER with the rest being borrowing costs). I’m hesitant to jump in at the moment since interest rates have been inching up, but I should probably just bite the bullet and stop trying to time the bond market :p
How do I go about executing bucket 3(i.e #4)? Can you provide some examples?
You’d probably need around 80% equities, 120% bonds, Implement with 80% equity ETF and the rest with Treasury futures.
I run something not quite at 80/120 but all these components thanks to ERN’s previous article. I’ll stick to that example though – and you definitely should check out that article. I hold the stock in physicals not futures. Buy the treasury futures with highest volume, which right now is March 21 – the market rolls over approx the last week before that month so you’ll be rolling forward in late Feb 21 to the June contracts. A 10-year futures contract requires about $2k in margin. Since I’m still accumulating I’m comfortable at 2x the margin requirement and if bond prices fall I replenish the margin with new funds (rebalancing).
In practice treasury futures contracts are large so the size of your portfolio dictates how close to 80/120 you can get with just the futures. One 10-year future is nominally about $140k right now so you’re looking at like 90-100k stock, 4k margin cash, and one futures contract. Given the gulf between contracts is quite large you can adjust by holding some bonds in physicals as you accumulate to the next 100k. Keep the 80% stock as a starting point. So if you have 150k right now, it’s around 120k stock, 26k physical bonds, 4k margin cash, one futures contract. One you approach 200k or so in physicals you switch over to two futures contracts.
Wow, thanks for sharing! Very important info! Good luck! 🙂
#9: buy and hold shares in great companies and avoid buying shares in crappy companies in crappy industries. that’s what we do in the malevolent missy stock series.
in all seriousness, what say you about something like motley fool stock advisor (my mentors) who have a time weighted return of 597% to 113% for the sp500 since 2002. 18 years is a pretty long track record. i have no affiliation except as a subscriber but with many good ideas from those cats my portfolio is up 284% to the vtsax return of 80% since 2016 (once again time weighted). i realize its “only” 5 years but it’s all i have to go on.
i also think if you’re buying steadily (like most real working people) vs. a one time only lump sum buy you can mix in some qqq or vgt with your vtsax to end up ahead of vtsax. anyhow, i always appreciate the thoughtful articles.
Stock picking? Not a fan. The market is smart enough to have already penalized the crappy companies.
Also, in the Motley Fool strategy, how much of that is backtested with hindsight bias? How much of that is lost to tax inefficiency due to trading in/out? How much is lost to trading costs/commissions?
How much more volatility would you have had from the MF portfolio? Can you share monthly returns for me to check?
Also, since 12/312002 the S&P 500 returned 500%. Probably more than the Motley portfolio after-tax. 🙂
But don’t get me wrong: What you are doing, intentionally or inadvertently, is likely a combination of the “high-beta” style and playing with the Fama-French SMB and HML factors. So, what you’re doing is not far from some of the things I’ve covered here.
ok, big ern. this is a long reply i’ll write a whole blog post for tomorrow on how we keep score. i hope you stop by from time to time in order to see if we’re still crushing the market over time or if we have crashed and burned!
we have absolutely gotten 3x the sp500 gains over the past 5 years with stock picking. should everyone do it? probably not. you really need the stomach to buy and hold through volatility, cut out losers and add to winners.
i don’t really know what you mean about hindsight bias or how to check that. i just printed up all 448 recommendations of which 244 are still open. there is not trading in/out to lost tax efficiency. most of the closed positions were either due to a good company being merged or taken private or losing bets being sold at a loss where you would save on taxes.. they tend to be patient (more than i used to be) and give their selections several years to make sure. also, trading costs are a thing of the past for normal people and a good online brokerage. even back when it was 10 bucks a trade 10 years ago an average of 12 closed trades/year wasn’t much of a hit. the rest if buy and hold and wait.
i have to correct you on the time weighted average scoring. i recalculated all the combined returns for open and closed rec’s just to make sure i have my facts straight. i realize that from the start of ’03 the spx is up 500% but they average every position as though when they initiated it they bought an equal amount of the index. if you include many of the closed positions where the comparative sp500 returns were low (compared to sometimes losing money for the individual selections) and average in the spx from the past couple of years (0-40% max) you end up with 113% average for the index with regard to time bought and held vs. any rec. i also re-calculated the average gain of the rec’s including the closed and losing positions and came up with ~530% average just like their claim. so i’ll stand by the fool’s approximate 5x outperformance claim. it helps they have 17 active rec’s with a return of over 2000% since the recommendation date. many of them were recommended more than once.
like i said before i’m not trying to convince any readers this is a way to go. i just present the facts of what i’ve seen and how our picks are performing. i put out our holdings each month for all the world to mock and we started the missy portfolio from scratch last year (we own all the missy stocks with real money) to show that well chosen stocks can beat the market. so far, so good. past performance is no guarantee of future results, as you know, but but our gains in the past 5 years are more than our best 5 years of 2 people working middle class jobs. let me know if you want more info to mess around with. peace.
I’m not saying that you did a bad job. Quite the opposite. As I said before, most of these recommendations are likely high-beta, Tech stock plays. IT did really well recently. Good for you. Is that really stock picking? Or sector picking you could have achieved with just a sector ETF without the hassle and the tax inefficiency? And again, simply beating the market might not mean that you’re a good stock picker (kinda, the punchline of my post, isn’t it?). If you had a beta of 1.5 and the market returns 10% p.a. and you make 15% p.a., then this difference can blow up to vast outperformance over time. And it’s not stock picking, but simply higher return for more market risk, thanks to the CAPM.
Also, the fact that you take a 500% return in the S&P and turn it into a 113% return demonstrates that your method of comparing returns active vs. passive is not proper. If you like, please provide your monthly portfolio returns. Once I see the monthly returns I can run some tests. There are standard, scientifically accepted tests to determine how much of your performance is due to a) market beta, b) style bias (value/growth, small/large, momentum, betting against beta), c) sector biases and d) stock-picking skills.
Believe me, I’ve some training in this (PhD, CFA) and some experience in this – running and testing active strategies while working as the chief econometrician for a large asset manager. And I have both confirmed and debunked people’s claims of their market-beating skills. 🙂
You don’t know what hindsight bias is? You answered it yourself:
And what a shock that their “track record” starts right after they killed their previous real money portfolio due to poor performance…
Uhm ,yeah, that’s another problem. Pick the right starting point right after the track record blew up in the dot-com bust. Oh, my!
Thanks so much for the reference. It’s in my bookmark list. 🙂
I would be very interested to see how one could combine “Tangency point plus leverage” into a “rising equity glidepath” post retirement to minimize safe withdrawal rate, particularly over a 60-year retirement.
Oops, I meant maximize safe withdrawal rate.
Good question. The glidepath would then not shift from S->B but keep the same S/B ratio and simply raise the risk level along the way. Not a bad idea. Have to think about how to implement this in the Google Sheet…
Hi ERN, I’m a long-time reader but only a rare poster. Great article! Quick question…..I was surprised by #3 above regarding the 70/30 stock/bond ratio being best for minimizing tail risk and maximizing SWR. From your SWR series, I’ve been looking to an 80/20 ratio as being better for a 30+ year retirement. Was I understanding that wrong, or is indeed 70/30 a better ratio from both a SORR and SWR perspective? If so, do you see that as a static ratio near/beyond retirement, or rather one that benefits from more of a glide path from 70/30 up to 80/20 (or beyond) after the main SORR years are past?
There is a very broad peak around the 60/40, 70/30 and 80/20 allocation most of the time. What exact allocation comes up as the best depends on the exact specification: when do you start the sims (1871 vs 1926), what’s the length of horizon (30y vs. 50y vs. 60y) and what’s your final value target (asset depletion vs. asset preservation, vs. partial preservation).
All depends on your personal preferences.
Right now I’m using call options in my IRA to simulate leverage. This is because there’s no tax drag on rolling over options since I don’t pay any taxes on transactions inside an IRA. I still own underlying stocks in my taxable account.
Another thing that might be worth considering is buying stocks that don’t pay dividends like BRK in your taxable account. It’s about just as good as the S&P 500 over a long time in terms of returns, but you don’t need to pay high taxes. When I was rebalancing (taking profits) I had to pay the top tax rate for capital gains so not getting dividends those years would have been much cheaper.
Yeah good point. You can also “beat” the market in the after-tax return sense. One way would be to split your investments into low-div/high-div and put them into the taxable/tax-advantaged accounts. I once proposed that:
I totally understand your view point, however I don’t even invest more than 50% in equities due to fear, thinking leverage is totally out of my “sleep well at night” approach. It suits more non risk-averse people for sure.
If that works for you, then you should stick with that plan. But over 10+ years you will come out ahead with 100% equities…
Let’s make a contract then. If in 10y the stock market is lower than today you pay me back the difference 😉
Sure. Though, I’m not going to agree to any bet where I have a non-zero probability of losing something but a zero probability of gaining anything. So, if you pay me the upside if the stock market outperforms, we can find a way to place a bet.
so you agree is non-zero probability of losing ! he got you there
It’s a probability game. Never claimed otherwise. He definitely dis not “get me there.”
Naive question perhaps – why we can’t there be a product available to us regular folks that replicates strategy 4? And if it really is higher reward and lower risk, why doesn’t everyone do it?
Could be regulatory. Everything that uses leverage is probably not deemed appropriate for retail clients.
But you’re right: one of the ETF providers should take a shot at this.
1) Leverage has a Russian roulette aspect to it. Is the portfolio levered during a major correction? If so, does continuing to be leveraged lead to a rebound, or is it retirement-over when the portfolio down a certain amount? See: https://www.bogleheads.org/forum/viewtopic.php?t=5934 for the epic story of a guy who went leveraged into 2008-09 and nearly lost everything.
2) When I look at the great pay-off-your-mortgage debate, I see that it is much easier to lower one’s WR if one pays off the mortgage, but by holding the mortgage one has more assets to invest in probably higher-yielding assets. If a SORR event is around the corner, one would do better paying off the mortgage. And because SORR events are *THE* thing we have to worry about, that’s the right answer. So as a rule, it is risky to take on a fixed obligation (a loan payment) that one expects to pay with variable revenue (portfolio earnings). Portfolio leveraging would increase fixed obligations – the obligation to buy an asset in the future – while increasing the variability of revenue, seemingly a step in the wrong direction for portfolio survival.
3) I’d be intrigued to see the nitty-gritty mechanics of how one takes a $1M portfolio value and determines which futures position to take to set the desired leveraged asset allocation. E.g. $1M stock exposure and $500k treasuries exposure. What exact contract do you enter? Does the position require maintenance?
4) I wonder if one could use margin to establish a costless collar position and harvest the higher likelihood of stocks going up than going down, all while minimizing maintenance margin and the odds of being liquidated near a bottom.
5) When even the actuaries / financial planner types who usually doubt the safety of 4% WRs are talking about the benefits of leverage, does that mean the top is in? 🙂
Great articale – which really is a synonym for your posts. One question thought: What about international diversification with FX gains attached? My dad tells me that when I was born the USD/CHF exchange rate was at 4.5 to 1. Now we’re at 0.9 to 1. OK, etf in ihr side of the pond are outragiously expensive at 10 BP, and probably a night-mare tax-wise, but still…
And I know that some argue US multinationals are globally diversified so one doesn’t need to look elsewhere. The same holds true for some European markets like DE or CH. Is there potential to beat US stocks or is it ‘our’ problem / view that the USD is loosing value because IT can, as the most liquid currency?
Sessions greetings – frohes Fest
Well, FX movements are mostly unpredictable, so they can go in your favor or against you.
There might be a case to be made that the USD is now on the decline again, so that would aid non-US stocks. But careful: non-US corporations will also lose a little bit in competitiveness through that.
Viele Gruesse und Frohe Weihnachten! 🙂
All good points.
1: don’t confuse a 2x equity portfolio (2x the risk) with a 2x max sSharpe Ratio (= same risk as 100% equities)
2: completely agree. Leverage is good when young and before retirement. Stay more cautious in retirement.
3: 1m portfolio, you’ll do $800k in equities and then keep $200 in mrgin cash and get exposure to $1.2m through Treasury futures. Pretty simple: divide $1.2m by the contract size (usually around 140k) and buy those contracts. Pretty simple to do. You have to roll the contracts every 3 months before the near-contract expires.
4: Not sure what “costless dollar” means. Free money is great. Not sure if anyone has figured that out. Maybe you have to become a politician? Or Defense contractor? Or lobbyist?
5: I have not heard actuaries/planners propose leverage in retirement.
This is a treasure trove of information. Thanks Big Ern for the early Christmas present! The option involving tangency point with leverage is very intriguing. I think I am going to try it with a small amount of leverage. I am already retired, so I am not looking for any where near a 2x leverage portfolio. My current portfolio is 60/40 stocks/bonds.
I am going to shift my asset allocation to 50/50 stocks/bonds. I will then add 25% 10-year treasury futures which will take my total portfolio to a 50/75 stock/bond split with I believe lies on the tangent line you mentioned that maximizes the Sharpe ratio and would give me a portfolio leverage of 1.25x.
I have access to futures trading in my Schwab brokerage account but I hadn’t used it. Till now I have only needed options trading access to implement your SPX put selling strategy. I looked into futures contracts today and – as best I can tell – IRS section 1256 tax treatment also applies to futures contracts (just like with index options). This is important because I can only trade futures in a brokerage account.
You mentioned that using treasury futures is a very cheap way to add leverage, so I looked into it. Schwab charges about $3.00 in fees per futures contract and treasury futures have a nominal value of $100,000 per contract, so rolling the contract quarterly would cost $24 per year (for a very reasonable expense ratio of 0.024%). Nice!
I read this excellent post several times, so I hope I am understanding how to implement your strategy #4. Thanks!
That’s how I would implement #4, correct.
And also correct, Treasury futures are 1256 contracts
Good luck! 🙂
Hi Big Ern,
I have now completed the transition from a 60/40 portfolio to the 50/75 portfolio using 10 year Treasury futures to add a bit of leverage. With the CAPE having just gone above 35 today, I feel a bit safer with a slightly lower allocation to equities.
Now that I am actually trading futures contracts, I am being exposed to investments I previously never even knew existed. For example, I just saw that I can invest in housing futures for specific cites such as LA, NY, DC that are traded through the Chicago Mercantile Exchange.
Then I remembered you are an accredited investor with private equity holdings in real estate. I also remembered you are not a fan of REITs. Have you ever looked into housing futures? It seems like a pretty tax efficient way to gain some exposure to the residential real estate market.
Housing futures are too exotic and illiquid. Nice idea. Poor practical solution, though.
I think I figured out why I had never heard of housing futures. They are very thinly traded – almost no volume at all. Most successful futures contracts are based on an underlying basket of highly liquid assets that are priced once every 15 seconds. Whereas private real estate indexes are calculated once a month or once a quarter. Basically, the housing futures market doesn’t work due to the illiquid nature of real estate and pricing based only on the most recently published stale index value.
Very true. I would not touch them. Too illiquid and exotic.
Thanks for walking me back from the ledge! I own a house, so I already have some residential real estate exposure. That’s good enough for me!
I am considering gradually shifting from 1.25x to 1.5x leverage (from 50/75 to 60/90) while maintaining the max-Sharpe ratio. That would seem similar to a glidepath approach, but I would be increasing risk through leverage adjustment rather than asset allocation shift. Which glidepath method do you think wins from a sequence risk perspective? Thanks!
I have simulated glidepaths without leverage. But I’ve never simulated a GP with the max-Sharpe portfolio + leverage.
Just bought my first treasury future contract on Schwab. Total fee was $2.27, so the expense ratio is even better than I thought.
Very cool! Congrats
A few questions on implementing the 80/120 portfolio.
1) Are we talking 10 year (/ZNH21 on TD Ameritrade)?
2) Do we buy ATM option contract?
3) Buy a few months out and then roll three weeks before expiration?
Interesting that someone at 80/20 allocation would just use futures to add another 100% treasures through option contract.
2) That’s an “option” (pardon the pun), but normally you’d implement this with just the futures, not options on futures.
3) yes. 1-3 weeks before expiration
Amazing and detailed post. This one is going to be bookmarked for me.
Great post! I really like this blog. Please, ERN, could you point me a book about the way you see investing?
100 baggers, the intelligent investor, 7 secrets to investing like warren buffet are a few, value investing is the way to go.
I don’t see ERN as a value investor…
Yeah, I’m not a big fan of value investing. Just stick with the index. It’s simpler and has been more profitable recently, 🙂
not sure my guy all investing should be value investing although it does depend on your risk tolerance I dont think anyone can say value investing doesnt work it just takes more dd, and a bit more risk depending on the type of value investing situation
boys go give my page a shout or a follow if thats possible going too be uploading post regularly on financial education management and literacy
Yeah, I might write a book about all my ideas at some point. But I have to write a book about safe withdrawal strategies first. 🙂
Great post, thanks for the information. I am trying to understand the chart in point #3 titled “30 Year Safe Withdrawal Stats 1920-Current”. My confusion is in the chart’s legend. The blue line is labeled Failsafe WR (right axis) yet the left axis is matching in blue color and labeled Failsafe Withdrawal Rate. The same goes for the gold line in the legend labeled as left axis even though the right axis is matching in color and label. Your charts are always fantastic so maybe I am just misunderstanding it?
Ooops, that’s a typo. In the legend the “right” vs. “left” has to be reversed. I will fix this now.
Thanks for the amazing post! But how do we actually get this higher return (or lower volatility)? From keeping 80/120 balance on every futures roll?
The implementation is up to the investor.
Example: 80% stock ETFs, 20% cash reserve and the 120% bond exposure through Treasury futures.
Since 2000 Small Cap Value (VSIAX) has nearly doubled the return of VTSAX. 597% vs 319%. It’s very easy to say, “well…that was a few years of a good run after tech crash in the early 2000’s.” Okay. lets start the clock at the beginning of 2004. From 2004 through the middle of 2019 the performance of the two asset classes are virtually identical. In the the last 18 months the S&P 10 (I say that jokingly because 10 stocks drove much of the returns of the SP500) have outperformed. Is that going to continue? I’ve got a strong bet that Small Cap Value is going to go another big run. We’ll see.
Yeah, and that’s an implressive return. But the outperformance is very episodic. It’s possible that this will continue. It’s possible that this will all reverse. My working assumption is that SCV will simply move according to the CAPM. Not more, not less.
Great article, as always.
You mention the advantages of homeownership but the problem of SORR for why no mortgage. But in the context of leverage to generate higher returns, I guess the question is: is there *any* level of absolute mortgage cost where you’d change?
I am now seeing 30 year fixed mortgages advertised at 2% (2.24% APR). Is that *still* not low enough to where a mortgage makes sense vs other forms of leverage for a retiree seeking higher returns for same risk level? Especially in a world of “helicopter money” making deflation all-but-impossible?
Depends on your stage in life. Young investors should leverage. Once retired, you should play safe.
Also: the low mortgage rate has to balanced against the extremely high CAPE. It’s probably a wash: lower mortgage rate but also much lower equity expected return.
I love the leverage one. These days, debt has gotten so much cheaper, robinhood lowered their interest rate from 5% to 2.5%, what! I would take any debt that offers me a 2.5% interest rate any day of the week and twice on sunday. And you are right, if you are young and you truly can afford to take those big risks like that, you can beat the stock market.
I thought this article might be an active investing strategy article but I am glad to see that it was not. Solid tips!
Thanks! And if you go the futures route, the implied borrowing rate is 0.1%! 🙂
A couple of thoughts:
1. It seems like value and low-beta stocks combined with leverage (to get risk back up, and return with it) might be a good strategy. Small-cap stocks already have high risk, so they wouldn’t need that.
2. You might write an article on rare events. For the longterm, I worry about the possibility of massive inflation, pandemic, punitive taxes by a Berniebro, etc. Do ordinary simulatoins cover thgat sort of thn g? Assign
Both good points, thank you!
One could also implement the low-beta style without leverage by simply using an 80/20 instead of 70/30 stock/bond portfolio.
I like the idea of the rare event study. Of course, the Great Depression is a rare event study of a quasi Bernie-Bro (=FDR), so if you’re safe during that event, you should probably make it through what the future might throw at us. 😉
Absolutely love this site and I thank you so much for all of the information you have provided. You have greatly expanded my knowledge of personal finance. Not sure if this is the correct place to post this (question is related to item 3) but I am hoping you could help me with something I am struggling with. I have a portfolio worth several million but am still several years from FIRE. My portfolio is very heavy equity index funds. I am struggling with how to de-risk. The 100% equity portfolio has come up quite a bit and has served me well (up to now). I had read about the glidepath and the go with 70% equities to raise the safe withdrawal rate. There is no way I can time the market to switch my portfolio right at the peak of the market. I believe you commented in one of your responses to start moving to a less equity heavy portfolio several years before FIRE. You have proven that without a doubt that less than 100% equity portfolio at retirement increases the SWR. What I am struggling with is if I switch to a 70/30 portfolio several years before a peak, I have a small portfolio vs. a 100% equity portfolio. How is that reflected in in your SWR analysis? If I did this switch before the peak, wouldn’t I have a larger SWR against a smaller portfolio?
Very good point. If you’re flexible with your retirement date you might as well keep the 100% until you retire. If you shift too early you might have a smaller nest egg (and thus the higher SWR in a smaller portfolio might be a wash, or worse).
Thanks a ton for your response. I am still struggling. If my goal is to maximize my safe withdrawal $ in retirement wouldn’t I be better off just keeping 100% equities all through retirement? (I know your answer is no based on your body of work and other responses but I cannot seem to tie it into this concept). On item 3 – the chart shows a SWR of ~ 3.75% for a 70% equity and a ~2.75% SWR rate for 100% equity but does this really tell the story? Wouldn’t another chart that shows the actual $ generated by a portfolio be more effective to demonstrate this? For example, a $2M portfolio at a market peak that is 70/30 would generated 75K (3.75% SWR) this compares to a 100% equity portfolio of $69K of an equivalent value of $2.5M and a SWR of 2.75% (this is made up math). The 100% equity portfolio would have benefitted from the run-up in the late stages of the market rally right before the peak thus generating a higher “effective” SWR. It feels like the bond shift does provide some downside protection at the very top of a peak (you have equated that to insurance and it will on average cost you) but it feels like expensive protection with a minor impact vs. just moving with the market. The protection feels like it would be fairly minimal (taking into account the lost portfolio value associated with the shift to bonds). What am I missing in my thought process?
Well, try it out. In the 30y case with 25% final value, the safe withdrawal amount was $77,255 with 75%/25%. But $58,681 with 100%/0%. So, 100% stocks will make your retirement better on average but worse during the worst case scenarios.
And to be pricise: what you’re missing in your thought process: the failsafe still occurs in Sep 1929 even with the OMY. So , you don’t benefit from the equity runup during the OMY. For such short periods, OMY or TMY, the months that generate the failsafe are still the same as in the baseline without OMY. So, under the failsafe, there is runup in the equity bull market you’d miss. 🙂
Nice work! If anyone is interested I’ve archived about 60 stock screens, good, and and ugly for everyone’s viewing pleasure!
Thanks for the link!
Great article. I would be curious to see your take on the leveraged etf NTSX by Wisdom Tree. It has a reasonable expense ratio compared to many etfs and seems to follow your prinicple of using leverage on the bond portion of the allocation. It is a 90/60 fund.
Yeah, I like it. Before you get to a large enough account size to trade futures yourself, the NTSX is definitely a nice way to get leveraged exposure.
Have you ever looked into accelerating dual momentum strategy as featured on ‘Engineered Portfolios’?
Sounds like it’s overfitted to perform well in past data. But then it has poor results if used going forward. I wonder how this process performed in 2020. Do you have the returns? Did they have to retroactively change their rules to make the 2020 returns look better?
Just one data point, but it looks like the strategy did well in 2020. If you look at their post https://engineeredportfolio.com/2018/05/02/accelerating-dual-momentum-investing/ under ‘How to “Live Trade” the Strategy?’ there’s a link to Portfolio Visualizer results, whose upper bound you can extend from 2018 to 2020.
It’s an interesting market timing strategy. I like to study the Stock vs. Bond part as part of the SWR study sometime in the future.
The original simple 12 month lookback has underperformed 11 of past 12 years, but what they call the 1+3+6 month score outperformed in 2020 but underperformed in 2019 and so far in 2021. As for the future, who knows!
Looking at the previous returns, you’re right it has slightly unperformed however I think the beauty of the strategy is it’s risk off trigger which does not expose you to the large occasional drawdowns
Exactly! That’s what I tried to convey with my other comment. You beat me to it! 🙂
Yes, I could see it being more useful in the withdrawal phase where large drawdowns are painful but its hard for me to endorse it during the contribution phase where drawdowns can even be beneficial if you have enough time for the recovery (10+ years) before retiring.
Completely agree! Don’t do momentum when accumulating!
It’s purely a hedge when I’m retirement!
“The original simple 12 month lookback has underperformed 11 of past 12 years”
Yes, because you had a few false alarms and got whipsawed. But that’s by design. You use the momentum signal to hedge against “the big one”, i.e., a repeat of the protracted drawdowns (1929-1930s, 1970s, 2001-3, 2007-9), which are the Sequence Risk retirement killers.
Hey ERN. Would you comment on implementing 80/120 with equity etf + treasury futures during times when the yield curve is inverted? Would it still make sense “in theory”? Intuitively it doesn’t because the treasury futures would have negative expected return, right?
Let’s start by first looking at the case when the yield curve is completely flat and the 10y yield = 3m yield.
As long as you have a negative S/B correlation you still have a use for bonds. Even with a slightly inverted YC you might still have some bonds in the max-Sharpe-Ratio portfolio.
Take a look at my classic post:
When you assume returns of 7.0%/1.5%/1.6% for stocks/bonds/cash and 15%/6% S/B risk and -0.3 correlation, you’ll still have 60/40 as the tangency portfolio.