Hi, I agree with you that everyone's risk appiette is different. But it isn't my intention to find a formula for customised situations.
The rule of thumb should be a rough gauge that is good enough for everyone, taking into account their age, and therefore appropriately assign risk.
The issue is really on the parameters surrounding the usage on the "rule of thumb" , which should still apply to all persons, generally.
OK, so I fundamentally take "View 2" in your numbering, which is basically to count it all, run whatever models you wish against the real and complete inputs, and then adjust the models and/or your decisions if you wish.
Let me start with this: I prefer a slightly different model for long-term investing. I prefer the "Vanguard Target Date" model, which is (oversimplifying slightly):
(a) Hold an 80-20 stocks/bonds split until 7 years before drawdown age, then
(b) Use the 7 year period before drawdown age to rebalance from 80-20 to 30-70, then
(c) Hold 30-70 thereafter, or possibly slowly rebalance to 20-80.
I prefer this model because retirement age expectations vary, life expectancies vary, and life expectancy is increasing (in Singapore anyway). The "110" makes assumptions about all of those factors, and there's no need to make those assumptions fixed since you (the long-term investor) can probably make some good estimates for yourself. It also implies a progressively increasing bond allocation through an entire working career, and I don't think cutting off that corner of the rectangle makes much sense. And 110-age implies more active trading because every year the percentages change, and I just don't think that's a good implication psychologically, at least. Prudent long-term investing should be very non-active, with an awful lot of "keep doing more the same way" doggedness. Vanguard fundamentally agrees with me because that's how they've designed their target date funds, so I'm in good company.
OK, here in Singapore there's a "problem": housing is just so damn expensive. If you consume your housing in owner-occupied form, then (for most households) the equity will eventually be a relatively huge share of total household wealth. That's just how it goes.
Now, we could just wave all that away and pretend that important wealth doesn't exist. But I don't like that idea. Doing that would mean, among other things, that an otherwise similarly situated family that buys a 4 bedroom condo compared to another family that buys a 2 bedroom HDB (3 room unit) would be taking equal total investment risks (in allocation percentage terms) if all other savings were invested in the same way. And that's just not right. There really is more stock-like risk for Family #1 compared to Family #2.
OK, let's suppose 50% of total household wealth is tied up in owner-occupied home equity, as in the scenarios you've outlined. And let's suppose we use the "Vanguard model" to evaluate the stock-bond portfolio split for a basic, point-in-time risk assessment. That'd mean that the other half (50%) would be split 60-40 between stocks and bonds. Is that pretty good? Yes, that looks OK to me as a rule of thumb.
OK, fast forward, and now we're getting near retirement. So the "Vanguard model" suggests a 30-70 split when the couple (let's suppose) hits age 65 (for example, as a classic/traditional retirement age). And let's suppose home equity represents 50% of total household wealth, still. Thus, even if the entire other half of total household wealth shifted to bonds, Vanguard's "rule of thumb" 30-70 split can never be achieved, not with this level of home equity. Is *that* a problem? Maybe, maybe not. At that point you can simply say, "Well, I'm not going to sell my home, and I still think it's the right size. And I feel comfortable with keeping 15% of my total wealth in stocks and the rest in bonds by the time I hit age 65." And that's fine, you can make that decision. As long as the home is not too burdensome in the circumstances, no problem. This signal just gives you something to think about, that's all. It's a good signal -- homes really are stupidly expensive in Singapore! -- and it's just reminding you of the importance to exercise caution in not getting too carried away with the expensive stuff.
Now, if you want to get slightly fancy, you could do something like this -- and it'd probably be consistent with the "Vanguard model" and its U.S. context. What you
could do is decide to re-run the model but exclude the portion of home equity that would equal a "right sized" HDB unit with a remaining leasehold to age 110. (Here we'll use that 110 age, with current understanding of life expectancies.) For example, if you're a couple at age 50 (or thereabouts), go look at the resale prices of 2 room or 3 room HDB units with 60 years of remaining leasehold. Whatever that reference price is, subtract it from the home equity you actually hold. Take the remaining home equity above that baseline -- call it "investment-oriented real estate equity" -- and run your stocks-bonds model based on that number.
That seems like a reasonable approach to me. You'd be taking the equity attributable to non-lavish, baseline housing needs out, which is really an expression of prepaid rent (60 years of remaining leasehold).
The "Vanguard model" has some implicit exclusions of necessity, in particular typical U.S. home equity (for an assumed single owner-occupied home that's non-lavish in the circumstances) and U.S. Social Security retirement benefits. And the percentages Vanguard applies to its target date funds might not be the right ones. The late Jack Bogle, Vanguard's founder, actually had a minor, polite disagreement with those who decided the target date fund allocations. Specifically, he didn't think they should have raised the fund percentage allocated to non-U.S. listed stocks. So far Bogle has been proven right, but maybe going forward he'll be proven wrong -- who knows.
Returning to the original point, though, I just think "signals" coming out of simple checks are useful, but maybe that's just me since I don't get worried about a signal that's flashing yellow. It's just something for me to consider, not necessarily or even very often for me to act on. Right now I happen to have a "signal" flashing that I'm notionally too bond heavy, as it happens. Yes, OK, but (upon investigation and further analysis) I have two reactions to that: (1) to some degree there's no remedy available, because I'm only one person in the household, and (2) I think I'm OK being a little more conservative at this time since I can see a couple large expenses that could pop up within the next few years. I'll keep an eye on the signal (and any others), then take another look about 6 months from now.
