Hi All,
I want to allocate a large portion of my portfolio (long term buy and hold) to capm based passive index fonds. The other part to Factor ETFs (next post). The simplest thing for the capm part would be to just put everything into an all-world, low cost like SPDR MSCI All Country World UCITS ETF (Acc) (0.12% TER).
Looking at the high P/E ratios of US stocks I feel some tilt towards non US-Stocks might be warranted. Thus splitting up into different ETFS (e.g. US, Ex-US and EM) and apply a custom allocation, deviating from actual market cap realities.
On the other hand I don’t really have a clear approach how to determine those custom allocation, which will make it hard to implement and rebalance such a tilt in the long run.
Any Thoughts, approaches you might want to share on this issue would be highly appreciated?
Note that Cederburg doesn’t even compare to a simple cap-weighted global allocation (besides what I find a kinda weird modelling). It always assumes some amount of domestic bias.
I used to have a custom allocation, personally I found it too hard to stick with it (it’s a bit painful to see years of underperformance and comparing yourself to VT) and switched to a single global ETF solution instead which is a lot easier to maintain.
Doesn’t the factor sleeve already do so? Looking at MSCI indexes Value, Momentum, and Small Cap are underweight USA. Quality is still overweight. For example, the MSCI ACWI Enhanced Value index only includes 20% US stocks.
Of course there are factor products that fix (or at least tether) country weights to those of a market cap weighted index. Notably AVWS.
Regardless, it makes some sense for now to split into US and ex-US (Dev-ex-US and EM) to get Swiss tax credit for the withholding taxes paid to the US. A problem that might get solved in the next few years (check: Reclaim withholding tax on Irish based ETF?).
I often feel that global cap weighted index is a sum of parts of regional cap weighted indexes. Maybe that’s why they just considered a typical investor who invests locally and also add international stocks
the VT & Chill crowd is rather small in the investor world but of course quite large on internet groups
You could do all kind of mental or statistic based exercises to come up with an allocation, but I doubt it will make a big difference in the long run whether you got a few percent more here or there. At least non that you could know about now.
So it basically could be based on gut feeling, and how many investment products you want to put up with.
You could have the World ETF and then add something else, or split it by region (I’m doing that) which might even have tiny advantages in fees and withholding taxes. Although for me, it’s mostly for my own entertainment and because I’m used to it from when I started.
There’s some studies how or why the US market (or the companies in there) became so successful, and some might argue it will continue to do so, but under CAPM logic that’s basically priced-in, already.
You could do 60% US, 20% Europe, 20% rest of world (mostly EM and Japan). Or 50% US, 30% Europe + extra CH, 20% rest of world. Or more EM or whatever. And so on. Either allocation may or may not do better then the other, but who knows?
Yeah it’s because their audience usually doesn’t even diversify beyond domestic. But still why not take global as a benchmark since that will benefit from whatever region is growing the most without biases.
I believe they do in the updated version. They also address many other questions. If you aren’t into reading, I recommend listening to the rational reminder podcast with him (the latest).
I’ve re-read/skimmed the new paper again, and I’m fairly sure a global market cap portfolio isn’t compared. (btw it’s interesting how removing US from the sample skews the results to be optimal with 45% domestic stock instead of 33%)
(It kinda makes sense they don’t include global market cap weighted because their academic novelty is in cleaning up data and building all the 10y blocks per country, they also mention that other papers often don’t even include foreign stock in the investment universe)
Looking at cederburgs findings it seems like trying to customize capm allocations beyond actual weights is not likely to outperform the market in the long run and there seems to be no systematic, proven way to do so anyways.
So i guess best for me is to put the capm part in one world etf and then pursue tilts in the multifactor part of the portfolio (about 30%) with etfs like avgs - mostly within small caps.
Any other recommendations on notable multifactor etfs you are using?
I actually decided to move away from US domiciled etfs (like voo, vt) and only use ucits compliant etfs. I might be missing out on chance to 15% dividend taxation benefits through da-1 but also also avoid any exposure to US estate tax. I understand this estate tax might no apply for swiss citizens up to a several mio threshold but still dont want my descendants to have to deal with the IRS..
„Using UCITS compliant etfs costs you 15% of dividend proceeds - not having to deal with the US IRS - priceless..“
One of the high-level conclusions was that the weight of the home country doesn’t matter that much on the low end (think, 0% is similar to 15%) but weights of 50% or more are detrimental, unless one’s home country is the US and the investor has a very strong belief that the US is exceptional.
Since we’re in Switzerland here, we can have a small home bias if we want, but it will likely be similar to a market weighed portfolio.
You could also think about the new Avantis ucits mutifactor etfs.
AVWC - msci world (no emerging markets) regional based, with moderate factor tilts.
AVEM - the emerging markets version of that (check that you have the ucits ticker when looking it up, the US version has the same ticker).
with those two you would basically have an all world moderately factor tilted version, that is still close to capm weighted, but still employs factors in a neat integrated package. The intergration part will help with internal trading efficiency and also goes over all capitalisations, helping to capture the premia where they appear.
That gives you an easy two fund portfolio.
Maybe add some home bias in i.e. SLICHA (therefore also reducing US exposure indirectly, and swiss stocks for swiss resident are very tax efficent due to no withholding tax whatsoever after reclaiming + small amount of tax free distributions), and you are set.
I strongly recommend everyone to consider “SPI extra” as the main component of the home bias. Not a financial advice, of course, but the top of SPI is (a) concentrated, and (b) gets lots of revenue outside of CH. I personally have both (7% for large, 8% for SPI extra).
Mit dem Lesen und der Teilnahme an diesem Forum bestätigst du, dass du die Forum-Richtlinien gelesen hast und damit einverstanden bist sowie den Haftungsausschluss auf http://www.mustachianpost.com/de/ akzeptierst.