Allocation rules are often necessary for disciplined portfolio management, but they create the possibility of committing an investor to a portfolio that's overly aggressive, overly conservative, or just dumb. The February 2018 market correction was an important test for my own allocation rules. Here I discuss the resulting changes I made to them, and how those changes helped prepare me for the bear market selloff in December.
I began using formal allocation rules in early 2017. They cover a wide range of factors including cash, bonds, value stocks, single stocks, and certain strategies within and across asset classes. I wanted the rules to be flexible enough to allow for things like buying dips in stock prices, so I originally designed formulas which adjusted the rules (one example being the minimum amount of cash I was required to hold) based on market conditions and stock valuation measures.
Since then I've come to three realizations: (1) it's hard to judge the reliability of allocation rules; (2) given the complex and often subjective nature of risk assessment, there's always room to question previous allocation decisions; and (3) it's at least as important for my rules to limit unnecessary trading as for them to be "right", and formulas which adjust things automatically can invite over-trading.
I've sought professional advice on my allocation rules, but the advisers I've talked to so far have shown little interest in them. Their focus is to see if my overall financial plan is sound at the top level, where boring stuff like life circumstances factor in (last I heard, I'm doing well at that level, though it took some fixing).
I originally put "Second-Level Rules" in the article title, since it's not my place to comment on the top level of a financial plan: I leave that to the advisers. I also emphasize that everything I write here is in the context of my own top-level plan, and your top-level plan may be quite different.
The second level I'm referring to (allocation rules) determines how much of my portfolio is invested in stocks, bonds, cash or alternative assets. In this article I summarize what's new and unchanged about these rules since the last time I discussed them, and I try to color the dry technical details with background on how the markets and my learning process informed each rule. I'm optimistic that the changes represent a refinement rather than a relaxation of standards. Although I've eased or removed certain requirements, I've put new ones in place which should pick up slack as needed.
Discussing alternatives before other asset classes may seem a bit backwards, but it'll simplify the rest of the article. Based on my definition of the term, alternative assets are 3% of my portfolio, of which 2.5% is precious metals and 0.6% is cryptocurrencies. I don't have any further crypto purchases planned, but I'll probably purchase more gold in the near future.
My allocation rules allow new purchases to increase my exposure to alternatives by at most 1.5 percentage points (1.5%) of total portfolio value per year (not counting income and capital appreciation from alternatives). Of the 1.5%, at least two thirds must be in the form of physical assets such as precious metals. For 2018 I've targeted a full increase from 3% to 4.5%, but I'm holding off on the actual purchases until there's a lull in stock volatility, as I can't resist prioritizing stock purchases at today's fire sale prices.
If I decide to increase the pace of alternatives purchases in a future year, I'll make the decision when the year arrives, and by at most 1% per year (i.e. possibly allowing 2.5% of purchases in 2019, and 3.5% in 2020, resulting in a 10.5% hypothetical allocation, give or take relative gains).
In past articles I mentioned a rule that limits my allocation to speculative assets, including precious metals and cryptocurrencies. That rule still applies, but I'm giving the alternative assets rule precedence over it: I'll increase the speculative allowance by a dollar for each dollar contribution allowed by the alternatives rule. In other words, I still consider precious metals speculative, but I also think I've underinvested in them. They remain an important means of diversification in an environment with serious long-term macro risks, such as high government debt, declining fertility rates and slowing labor productivity growth.
Cash, Bonds and Stocks
One thing I don't plan to change is treating cash and bonds equally under a single rule, since I view both as having similar risk/return characteristics (especially since I keep much of my cash in a money market fund, giving it aspects of a fixed income investment). While I overwhelmingly prefer cash over bonds for liquidity reasons, I otherwise have little to say about the merits of one versus the other.
I make a distinction between personal emergency cash (which I do not include in any of the portfolio breakdowns I post online, nor do I discuss my process for managing it) and cash in my investment portfolio. Outside this paragraph, "cash" refers only to investment cash.
As I mentioned, I initially built a formula which automatically adjusted my allocation target for cash and bonds based on a few macroeconomic and stock valuation metrics. I didn't mind the 16% minimum exposure the formula required at the end of 2017 (in fact, I had a whopping 21% exposure to cash and bonds at the time), but when it only dipped to 14% at the trough of the February selloff, I began to suspect it was too conservative.
The biggest change in my attitude toward cash and bonds over the past year came when I acknowledged that stock selloffs have historically been minuscule compared to the gains in-between them. Even the era-defining crash of 2008 is dwarfed in magnitude by the gains in U.S. indices since then. While I still see cash and bonds as an important diversification tool, I believe I overestimated the merits of holding cash for dip-buying.
I've replaced the aforementioned formula with a set of conditions for making incremental tweaks to the cash and bonds rule over time. The new approach acknowledges my inability to anticipate every allocation concern, while limiting the magnitude of changes which may prove ill-conceived later on. It should also encourage less trading than the formula did.
I allow two types of adjustments, largely at my discretion, which are detailed below. All allocation percentages here apply only to the stocks, bonds and cash in my portfolio, which I call "traditional assets" for convenience (alternative assets do not figure into the following percentages). Also, the allocation percentages for cash and bonds are minimum requirements for those assets.
The Selloff Adjustment
While I acknowledge that holding cash for dip-buying is overrated, I'm still torn between the merits of cash for diversification and my belief that the value of cash lies mainly in the purchases that can be made with it, considering historical inflation. Therefore, I still have a dip-buying mechanism, but one which should result in less stockpiling of cash than in the past, since I've built in a lower long-term minimum for cash and bonds (10% of traditional assets).
During a stock selloff, I may reduce my minimum target for cash and bonds by up to 3 percentage points (3%) of traditional assets for each 4.5% the S&P 500 goes below its most recent bull market peak. Suppose, for example, that my traditional assets are 85% stock, 15% cash, and the S&P sells off by 4.5%. In that case, I could change my allocation to 88% stock and 12% cash. If the index were to continue down to 13.5% below its peak, I could adjust again to 94% stock and 6% cash.
Highly observant individuals will note here that some extra conditions are needed to prevent market volatility from allowing duplicate selloff adjustments (i.e. "The market bounced above my threshold, and now it's back below it. Time for more buying!"). To address this, I set waiting periods in-between each adjustment. If I adjust for a 4.5% selloff, I'll wait 9 months before doing so again. If the S&P continues into a 9% selloff, and I adjust again, I'll add 9 months to the first adjustment's waiting period, then set an 18-month period for the second adjustment. The waiting periods continue to increase in this fashion, by 9 months for each additional 4.5% of selloff magnitude.
Based on the February 2018 selloff, I currently must wait until February 2021 before buying into any selloff up to 9% below peak. In addition, due to the December 2018 selloff, I must wait until December 2021 before buying into any selloff of 13.5% or 18%. Obviously I'm expecting a lot of patience from myself, but based on the historical frequency of selloffs of these magnitudes, I also expect that buying opportunities like December 2018 will only happen every three or more years on average.
The Quarterly Adjustment
This adjustment comes in a few forms. If cash and bonds are over 10% of my traditional assets (the long-term target I mentioned earlier), I may either increase or reduce them by 1% (i.e. from 13% to 12%, as I did in July 2018).
If cash and bonds are below 10%, however, I cannot use the quarterly adjustment to decrease them any further. If I've already used a selloff adjustment to take cash and bonds below 10% (as I did in December), I may leave that allocation in place only until the S&P 500 hits a new record high. Once that happens, I must use the quarterly adjustment to increase cash and bonds by 1% in each quarter that the S&P finishes above the previous peak I traded on, until they're back at 10% of traditional assets.
Neither of the above adjustments may be deferred and "carried over" to the future if I choose not to make them during the opportunities provided. Also, since I'm defining only minimum amounts for cash and bonds, I may exceed the targets in a variety of ways, including wage income, dividend income, asset sales and fund distributions.
As of December 31, 2018, my target for cash and bonds is 5% of traditional assets, down from 13% when I implemented these rules in April. I got from 13% to 11% with two quarterly adjustments in July and October, then from 11% to 5% with two selloff adjustments in December. I tweeted about each of these adjustments as I made them, and details on the stock purchases involved can be found in the Twitter feed linked to my profile.
To illustrate how I'll use these rules to build back my cash and bond position: suppose the S&P found its true bottom on December 24, and that it will return to record highs in December 2019, never to go back below 2,930. In that case, my rules require five quarterly adjustments of 1%, starting in January 2020 and ending in January 2021. After the 2021 adjustment, I would be back at 10% cash and bonds.
I hope this article proves useful to others looking for a way to conceptualize an allocation plan which offers modest but not excessive flexibility. It may be hard to believe, but when I was drafting this version of my rules, I thought I had greatly simplified things by focusing on the allowed increments of change while dealing less with the inputs deciding the change. My experience suggests that any allocation plan which goes beyond a single, permanent ratio (i.e. "90% stocks, 10% cash, end of story") is inherently complicated, especially with a meaningful dip-buying mechanism built in. So far, I'm satisfied with this plan -- especially since it set me up for plenty of buying in December.
As the title implies, I plan to follow up with a second article which covers my allocation rules within and across the aforementioned asset classes: value versus growth, domestic versus international, individual issues versus funds, and so forth. I hope to avoid writing a book in the process.
Disclosure: I am/we are long ALL ASSETS, UNLESS OTHERWISE STATED. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.