New 3a solution from Finpension

The difference is because shows the current status (17.12.2020), while the one in your picture, given in the factsheet from finpension website, is the snapshot on 31.10.2020:

Regarding your 2nd question, then I have to say I am very new at this, so I cannot really tell what kind of strategy each of those funds serves.

1 Like

Hey, great job, quite helpful! I have a few questions though:

  • In 1) Country replication, why do SMI and SPI Extra indexes are not 100% Switzerland Shares (you use 94 + 6, but both should be 100% Swiss)? Did you try to add already some cap size on the country match criteria?
  • Same question for 3) Final replication (there you have 75+25)

What I did extra is to reiterate the numbers chosen to be entered in finpension and retrofit them into sheets 1 and 2, until I got a closer match on both Country and Size. I then end up with:

73% CSIF (CH) III Equity World ex CH Blue - Pension Fund ZB
09% CSIF (CH) III Equity World ex CH Small Cap Blue - Pension Fund DB
14% CSIF (CH) Equity Emerging Markets Blue DB
02% CSIF (CH) Equity Switzerland Large Cap Blue ZB
01% CSIF (CH) Equity Switzerland Small & Mid Cap ZB

Another thing: I wonder now if we should use the “Quality” version of the World ex CH fund, as it seems to pay-off despite the higher TER.

1 Like

That would be factor investing. Nothing wrong with it, but you need to commit.

Be aware: past =\ future. Performance chasing is a common mistake.

1 Like

SMI + SPI Extra represents the Swiss stock market without overlap. So I tried to use the cap size based on I was expecting more something like 80% / 20%, but if I use the correct cap size value it’s 94/6.

Corrected now. It did not affect the formulas as they were linked directly to sheet 1).

Exactly it’s iterative. In the exemple I have currently on the sheet you see that by matching mid-cap, there is overall less US exposure, and over-representation of small caps. So by doing the iteration you arrive again at better US exposure (e.g. country match) and probably small-cap under-representation. I guess you have now under-exposure of mid-cap, as it is not possible to have a perfect match. That’s why sheet 2) is really useful to understand the match, e.g. the choices you have to make.

Fully agree, and VT should be the way to go for passive investing. For the moment I just replicate VT because I’m starting to invest only now in VT on IB. Maybe when I have a lot of VT in IB I might diversify a small part of the portfolio which could be the part in finpension, as this Quality fund is accessible only there and not on IB if I got it well (only for institutional investors).

1 Like

Note also it’s not perfect because the size / type matrices do not sum to 100% in the databases, so I renormalized to 100…

1 Like

I did a detailed comparison here of the different platforms:


There were a few very “creative” approaches mentioned on how to replicate VT. They lead to fairly wrong outcomes; so lets take a fact based approach to this:

  • Both FTSE All World and MSCI ACWI IMI try to capture the entire markets; their market cap differs a bit as they use different Minimum criteria (e.g. market liquidity) but fundamentally, we can conclude that FTSE All World equates to MSCI ACWI IMI
  • MSCI ACWI IMI consists of (USD Market Cap as per 30 Nov '20; as per MSCI Index Fact-Sheets):
    – MSCI World: 49’095’177.07
    – MSCI World Small Cap: 6’676’651.64
    – MSCI Emerging Markets: 7’116’068.89
    – MSCI Emerging Markets Small Cap: 788’027.40
  • We further have MSCI Switzerland IMI, that consists of (USD Market Cap as per 30 Nov '20)
    – MSCI Switzerland IMI: 1’605’040.50
    – MSCI Switzerland: 1’468’492.60
    – MSCI Switzerland Small Cap: 1’360’547.90

Given that FinPension works with World ex CHF Funds, we may conclude that a Total Market Cap can be achieved with:

  • 75.7% MSCI World ex CH (Market Cap of 47’626’684.47; MSCI World minus MSCI Switzerland)
  • 10.4% MSCI World Small Cap (Market Cap of 6’540’103’74; MSCI World Small Cap minus MSCI Switzerland Small Cap)
  • 11.3% MSCI Emerging Markets (Market Cap of 7’116.068.8 as per the above)
    => We ignore MSCI Emerging Markes Small Cap as this was tiny (788’027.40); one could as well either shift this to MSCI World Small Cap (my recommendation) or to MSCI Emerging Markets
  • 2.6% MSCI Switzerland IMI (Market Cap of 1’605’040-50 as per the above)

When we now have a look at MSCI Switzerland, we can break this down as follows (as per 31 Dec '18):

  • SPI: 1’197b
  • SMI Traditional: 974b
  • SMI: 776b (the capping of the top holding “removes” some covered Market Cap
  • SPI Extra: 223b

Given that FinPension foresees we invest in SMI and SPI Extra Index Funds, we therefore break the Swiss Part down as follows:

  • 77.7% SMI: 776b
  • 22.3% SPI Extra: 223b

In total, we end up with (assuming 100% investment):

  • 75.7% MSCI World ex CH
  • 10.4% MSCI World Small Cap ex CH
  • 11.3% MSCI Emerging Markets ex CH
  • 2% SMI
  • 0.6% SPI Extra
    This is the closest to VT that you can achieve with MSCI based Index Fonds

75% MSCI World ex CH
10% MSCI World Small Cap ex CH
11% MSCI Emerging Markets
2% SMI
1% SPI Extra
1% cash

And then disable rebalancing (if it’s possible in 3a).


Thanks for your work @Cortana and @TeaCup as well as the others above. One naive question, why would you disable rebalancing? The funds on finpension are in CHF. Doesn’t that mean, that there are no rebalancing costs occurring?

I guess the idea is that since it’s already cap weighted no rebalancing should be necessary. (Not sure if it’s true due to accumulated dividends that might make things diverge).

What I’m wondering is whether it makes sense to include anything with less than 5%, isn’t it overkill to replicate in that much detail? If it becomes a large fraction of the equity universe, you’d only miss a tiny bit anyway, right? (basically you’d miss that part growing from 2% to 5% so this is somewhat bounded).
Am I overlooking something? This would make it much simpler to just ignore SMI/SPI.

1 Like

Noob question : I have CSIF but no MSCI in finpension…?

CSIF are product names, MSCI the corresponding index names…

Screenshot of the product names:

Screenshot of the above products with their corresponding index names:
Screenshot 2020-12-22 at 14.00.39

1 Like

Not possible that why I recommended to remove the CH part

1 Like

Thank you very much, more clear that way.

Because these fixed proportions are only true now. In 3 years MSCI Emerging Markets could be at 13% instead of 10% for example, so you don’t want Finpension to rebalance back to 10%. You let it be so it stays at market cap weights. You could also adjust these %-settings on a regular basis, but that’s too much work IMO.

As @wapiti pointed out, disabling rebalancing doesn’t work. What are the rebalancing bands of Finpension? Viac uses 2% absolute. If it’s the same at Finpension, I wouldn’t remove Switzerland. There won’t be a rebalancing with Swiss assets as there is no way that SMI could grow from 2% to 4% or SPI Extra from 1% to 3% to cause rebalancing problems.

1 Like

“If the weighting of a fund deviates from the target weighting by more than one percentage point, your entire portfolio will be realigned.”

The easiest would be to chose 99% MSCI World ex CH, but you miss small-cap and EM, otherwise you need to adjust on a regular basis. Each buy and sell has a cost of around 0.03% to 0.10% (displayed on the factsheet). Finpension like Viac use netting to reduce this cost.

“But before we do so, we check whether other customers within the finpension 3a Retirement Savings Foundation client base have orders to the contrary. In the end, we only trade what is necessary below the line. The rest is charged internally at the currently valid price. Fund units are moved from client X, who wishes to sell, to client Y, who wishes to buy the same fund units. Of course, this internal settlement, which is called netting, is fully automated.”

I would also recommend to take into account your whole assets. Because if 3a is 5% of your total asset, the impact of your 3a strategy will be small.

1 Like

There is a theory that says that in the long-run real interest rates are the same for all currencies. Then, the loss of value of the dollar over 20 years is partly compensated by the higher interest rate, due to the inflation difference. If the loss of value corresponds to the difference of interest rates and to the difference of inflation, it does not matter which currency you hold, and hedging cannot help make a difference here.

But if the loss of value was larger than the difference of interest rates, then hedging would have helped.

If you take interest rates into account, then USD compared to CHF has not dropped as much as the impression given by the graph in 20 years. The graph would have to be redrawn inflation-corrected to really see the drop that hedging would protect against.


This makes perfect sense, thanks a lot for the detailed explanation :pray:

1 Like

The quality fund contains only companies that have " historically high return on
equity, stable year-on-year growth, and a low leverage ratio"; the world ex ch fund is just a basket of companies with no consideration at all for growth, ROE, or debt of the individual companies in the fund. Performance chasing is a common mistake - if you are chosing the quality fund only based only on the fact that it outperformed the world fund since 2015, then thats the wrong reason; on the other hand if you are chosing the quality fund becuase you believe that companies with high ROI, conistent growth, and low debt will outperform the general world market over the long run, then that is a good reason to chose the quality fund. In my opinion it is unlikely that the world fund will ever outperform the quality fund over the long term; but that’s just my opinion and of course i could be wrong.

1 Like

You need to be aware that it is always relative to the price.

So why would the market value two companies with the same profitability different?

By reading and partipating to this forum, you confirm you have read and agree with the disclaimer presented on
En lisant et participant à ce forum, vous confirmez avoir lu et être d'accord avec l'avis de dégagement de responsabilité présenté sur