Throw It All Out And Start Over?

There was a lot of content from various places over the weekend about whether it is time to go back into bonds, what retired investors should do for yield and even whether retirees are better off going 100% into equities. 

We can start with Barron's which suggests it is time to "buy bonds" as the "60/40 bounces back." Barron's talks some about safe withdrawal rates and the yields now available. A little more specifically the need for diversified portfolios persists with the implication that bonds are the way to get this done. John Renkenthaler from Morningstar is quoted "the 60/40 remains attractive for the long term."

This chart contributes to the logic supporting a 60/40 portfolio. The blue line is equities and the red line is bonds up until the end of 2020.

That's a pretty smooth, uninterrupted, multi-decade run for bonds. Since it only goes back to 1988 it doesn't track from the highest interest rates from earlier in the 80's, the ten year was at 8.5% when the chart starts, down to 1.09% per Yahoo, after bottoming out in the neighborhood of 0.50%. As a matter of math, it cannot repeat the run from 8.5% down to 0.50% let alone from the all time high of 15% down to 0.50%. 

Barron's also noted that 60/40 was up 9.6% in November. That's great but as a repeat from recent posts, bonds with any sort of duration have taken on equity beta. Look at a chart, that will tell you. The 9.6% for 60/40 in November, the TLT ETF was up 9.5%, tells you bonds with any sort of duration have taken on equity beta. Do you want that kind of volatility from the fixed income in your portfolio? There's no wrong answer, do you want that kind of volatility? The working answer here has been no. The one way trade is over, bonds with duration now trade in both directions and do so with a lot of volatility.

It's not that 60/40, or some other combination of numbers is bad or dead, more like how we build the 40 or other number maybe needs to be different. That's not a new idea here by any means. Here's an idea though that might be new that comes from a study titled Beyond The Status Quo:A Critical Assessment Of Lifecycle Investment Advice which after running a Monte Carlo Simulation concludes that investors should not have any bonds but instead allocate 50% to domestic equities and 50% to international equities. They say that mix will yield a better result than a portfolio that diversifies across stocks and bonds.

It's hard to model the idea and get the same result with mutual funds on Portfoliovisualizer. I was unable to do so but the idea is simple, you can decide if it is correct enough to implement for yourself or not. Between domestic and foreign equities one must out perform the other. In the 90's and then again in the last 12 years or so, domestic has outperformed foreign by a mile. In the 2000's, foreign outperformed and a large allocation to foreign was very important during that time. The study concludes that the zigzag effect between domestic and foreign equities leads to a better long term result than the zigzag between domestic equities and bonds, a zigzag effect that I've been arguing is dramatically less beneficial for investors than it used to be.

Bloomberg dove into the study. This article seemed to be a mix of explaining the Lifecycle study and investment professionals defending allocations to bonds using various logic. One quote in defense really stuck out. Joe Quinlan from Merrill Lynch believes the "stability" that bonds offer "allow clients to sleep at night."

This chart would seem to refute the stability factor.

The actual report was 70 pages so I did not read the whole thing. I did a few control-f search through the document and found no mention of things like sequence of return risk or cash management strategies. 

Using Vanguard funds as proxies, VFINX for domestic equities and VGTSX for international equities, an investor who retired with $1 million on Dec 31, 2007 who immediately took out their 4% for 2008 would have had $576,000 at the end of 2008. That combo of funds to align with the Lifecycle study would pretty much have been blown up. That's probably the worst scenario, based on real market results, we could possibly concoct. If the person didn't take anymore money out, I realize that's not too realistic, the account would have been back to breakeven five and half years later. If they continued to take $40,000 out per year they would have briefly gotten back to their highwater mark in early 2021 but would today be down slightly from their $1 million starting point.

A similar story with only a 20% drop in that first year would still be bad. 

I think the companion to the Lifecycle study, all-equity allocation has to be a cash bucket for sequence of return risk. Bear markets typically range from 18-30 months in duration, so that can be a decent guide to decide how much to set aside. I think two years worth of expected expenses, insulated from market volatility in this context is a good generic benchmark. 

In past posts we've looked at labeling asset classes with more descriptive names as opposed to just the nouns that they are like stocks and bonds. The Quinlan quote above gives us a good descriptor. Instead of bonds, maybe it would be better to think in terms of allocating to stability to help manage normal, equity market volatility. 

No one has ever pushed back when I say that I think what most investors want out of fixed income is yield with very little volatility. Assuming that is generally correct then maybe instead of stocks and bonds, a simple allocation strategy could be to allocate to growth and stability. Growth could take in anything that has equity beta which could include several fixed income products like long dated treasuries and as we looked at the other day, convertible securities. Just because long dated treasuries might fall under this description does not mean I want to own them, I don't and the jury is still out for me on convertibles, they're just examples. 

There are countless ways to build a stability allocation that we write about all the time including short term fixed income, there are some newer ETFs that appear to offer stability, at least they target stability, like BOXX, JAAA (client holding) and a very new one SOF. There are a lot of arbitrage funds, both merger and convertible that are very stable. Convertible arb goes long convertible securities and short the common with a result that looks nothing like just buying a convertible bond fund. There are a lot of other types of alts that fit the bill of being horizontal lines that tilt upward as I've described them before. Some pay yield and some inch up very slowly over time, either way these are all pretty stable the vast majority of the time if not always. 

Portfoliovisualizer appears to have done an upgrade so there's a lot of information there. Portfolio 1 is fairly apples to apples versus VBAIX using arbitrage alts instead of bonds, standard deviation comes down a little with a noticeable bump to the CAGR, importantly to me is that Portfolio 1 was better than VBAIX by over 500 basis points in 2022. If longer bonds end up being the roller coaster that I think then the comparison might favor Portfolio 1 by a wider margin in a couple of years. If not, if rates go down and stay down then yes, Portfolio 1 would lag but the premise is preferring to avoid volatility from what traditionally is the fixed income allocation. 

Portfolio 2 above has 65% in growth. I wanted to see if the stability exposure could offset enough growth volatility to allow for that extra growth potential. Technically, no it doesn't. The volatility isn't insanely high or anything but it doesn't work as true leveraging down as I've described in other posts. 

To get the leveraging down effect we should add another descriptive asset class. I would leave growth, which is synonymous with equities, the same, some sort of "normal" allocation and make room by reducing some from stability to create anti-beta or we could call it negative correlation, doesn't much matter other than anti-beta has fewer syllables. The point is this is where you'd add negatively correlated assets which provide more diversification benefit pound for pound than stability. 

I added 5% in BTAL (client and personal holding) to Portfolio's 1 and 2, taking from the stability sleeve. The 5% allocation to anti-beta is pretty much just a tweak but as I often say, a little bit goes a long way. It is still a market portfolio with a "normal" allocation to growth but in the case of Portfolio 2 and it's 65% allocation to growth, it results in a higher CAGR and lower standard deviation than VBAIX. The standard deviation is only modestly lower, yes but the CAGR is significantly higher. The offset from anti-beta allows the portfolio to have more growth opportunity and again, much less pain in 2022.

This was a long post but we really dug into some good detail on portfolio construction. 

The information, analysis and opinions expressed herein reflect our judgment and opinions as of the date of writing and are subject to change at any time without notice. They are not intended to constitute legal, tax, securities or investment advice or a recommended course of action in any given situation.