Equal weight equivalent of VWRL (or similar)

[hiriskpaul]

The US market has equal weight ETFs, but you cannot buy them very easily anymore, or any other US listed ETFS.

Larry Swedroe on equal weight ETFs:

https://www.etf.com/sections/index-inve ... nopaging=1

In brief, you can get similar returns at lower cost by investing say 80% in a whole of market cap weighted fund and 10% each in small cap and value ETFs.

[Lootman]

The US market has equal weight ETFs, but you cannot buy them very easily anymore, or any other US listed ETFS.

Not a problem if you already own them however. I only have one but it has not been affected by this change.

Shame really because the OP could achieve his aim using the wide-range of single country ETFs listed in the US. He could equal-weight the US and Peru, if he wanted to. Or Egypt and Japan. Or Malaysia and Germany.

Of course you could always use a share-dealing account domiciled elsewhere. Does the Isle of Man have this restriction? Luxembourg? Gibraltar?

[dspp]

Hi GeoffF100.

Are you questioning value investing in general (in which case, probably best cover it in another thread elsewhere) or suggesting that the loose analogy I drew is way off base?

Regards, Newroad

Your summary, of:

Cap weighted tracker ETF(s) and similar: passive momentum investing
Equal weighted tracker ETF(s) and similar: passive value investing

is correct to my mind. Whether it is biased towards or against value vs momentum would very much depend on the: sequence of returns; the general market trend; and the duration. The numbers you have shown for 30-year horizons tend to favour value in the last 90-years. Whether that holds true for all time is yet to be seen. The way I read Bogle is, he thinks trying to get too clever is likely to end in tears.

regards, dspp

[GeoffF100]

Hi GeoffF100.

Are you questioning value investing in general (in which case, probably best cover it in another thread elsewhere) or suggesting that the loose analogy I drew is way off base?

Regards, Newroad

I am indeed questioning value investing, as did Jack Bogle. If you invest in a value index, you are investing in a subset of the stocks, and will have very little or nothing in some sectors. That reduces diversification and increases risk. Value stocks themselves have out performed over the very long term (mostly before their out performance was noticed), but also have higher risk. The safest approach (that can accommodate >50% of the market) is to invest in a market weighted tracker as Bogle advocated. If you want to do something else in the hope of achieve higher risk, that is up to you, but it is active investing.

[1nvest]

The safest approach (that can accommodate >50% of the market) is to invest in a market weighted tracker as Bogle advocated. If you want to do something else in the hope of achieve higher risk, that is up to you, but it is active investing.

Bogle's preferred choice was to initially equal capital weight the largest 50 stocks, buy and hold. No reason why that couldn't be supported by >50% of the market as all that occurs is that each case finds its own natural cap weighting over time.

Buying into what has drifted to find its own cap weighting (current/existing Index) is a momentum following stance. Allocates more to the stocks/sectors that have performed well, which is to opine that the historic better holdings will continue to either outperform or at least pace the average. In some cases over-stretched stocks/sectors will mean-revert.

Each day you are in the market is no different to lumping all-in on that day. Rather than adopting a initial capital equal weighting and leaving that to run, you could opt to periodically end/restart all-over, which has the effect of resetting back to equal weightings and reducing the risk of over-concentration into a single stock/sector. But depending upon the frequency could be cost/tax prohibitive. A partial more cost/tax efficient choice is somewhere between the extremes of rebalancing back to equal weights continually, or permanently leaving as-is (buy and hold), such as the top-slicing method that Terry (TJH) uses with his HYP (reduce exposure to the holdings that have relatively outperformed).

[Newroad]

Hi GeoffF100.

So you are both questioning value investing in general AND disagreeing with my loose analogies. I'm not going to get into the first - that can be done somewhere else - and almost certainly already has been.

Others (1nvest, dspp) appear to think my characterisations were reasonable - I also remain convinced of this. An equal weight ETF is not "active investing" in that you are not choosing which stocks to invest in, other than selecting the index or whatever to follow and whether you choose to do it completely or via sampling (both of which are similar considerations in capitalisation weighted ETF's).

So, if you choose to do this by capitalisation weighting, you are relatively over-weighting stocks which have grown and/or are growing - momentum investing, more or less.

Conversely, if you choose to do this by equal weighting, you are playing a relatively more mean reversion, small cap strategy - closer to value investing.

Regards, Newroad

[hiriskpaul]

I have just used portfolio visualizer to look at the performance since launch of RSP, which I think is the longest running equal weight ETF, compared to the total (US) stock market ETF VIX and Vanguard small cap index fund NAESX. I used NAESX instead of the ETF as it provided a longer history.

Overall performance was very close, with the small cap fund slightly ahead at 9.94% CAGR, compared with 9.88% VIX, 9.75% NAESX. However, the average hides a lot of detail! Up to the end of 2019 the differences were larger and RSP was ahead of VTI (10.82% NAESX, 10.55% RSP, 10.11% VTI). This year, as with the major drawdown in the GFC showed VTI to be more resilient than the other funds and allowed is to catch up. From a risk perspective, visible in volatility and absolute drawdowns, NAESX is more risky than RSP and RSP more risky than VTI. No surprises there. On a risk adjusted basis, VTI comes out the best on both Sharpe and Sortino. Sharpe (0.63 VTI, 0.57 RSP, 0.53 NAESX), Sortino (0.92 VTI, 0.83 RSP, 0.77 NAESX).

Comparing 50/50 VTI/NAESX (annually rebalanced) to RSP, the lines on the chart practically sit on top of each other, but with VTI/NAESX just slightly ahead on return, risk and risk adjusted return.

Based on these results I really cannot say I miss equal weighted ETFs.

Off topic perhaps, but for something different which totally hammers cap weighted trackers on risk adjusted returns, you might want to consider minimum volatility ETFs. One I hold is USMV, unfortunately no longer available for UK retail investors, but LSE listed ETFs are now available (with higher fees!). Since launch in 2011 USMV as matched VTI for returns CAGR (14.36% VTI, 14.34% USMV), but shows lower volatility and drawdowns resulting in much better risk adjusted returns. Sharpe (1.20 VTI, 1.56 USMV), Sortino (1.96 VTI, 2.94 USMV).


ps I reject the idea that cap weighted funds are in any way momentum funds. They are market neutral, which is the whole point of them.

[hiriskpaul]

Hi GeoffF100.

So you are both questioning value investing in general AND disagreeing with my loose analogies. I'm not going to get into the first - that can be done somewhere else - and almost certainly already has been.

Others (1nvest, dspp) appear to think my characterisations were reasonable - I also remain convinced of this. An equal weight ETF is not "active investing" in that you are not choosing which stocks to invest in, other than selecting the index or whatever to follow and whether you choose to do it completely or via sampling (both of which are similar considerations in capitalisation weighted ETF's).

So, if you choose to do this by capitalisation weighting, you are relatively over-weighting stocks which have grown and/or are growing - momentum investing, more or less.

Conversely, if you choose to do this by equal weighting, you are playing a relatively more mean reversion, small cap strategy - closer to value investing.

Regards, Newroad

I have to say there are a lot of problems with equal weight investing:

- RSP chooses to equal weight the largest 500 US companies. But if we are equal weighting, why choose those? What about the next 500 companies which have achieved zero weighting according to this scheme? Cap weighting avoids this irrationality by setting the weight according to the size of the company so beyond a certain point excluding the tiniest companies makes little difference to returns compared to the market.
- A company splits in 2 and if both end up within the selection criteria, you have to double up. If both fall outside the criteria, you sell both. How does this make any rational sense?
- On rebalancing, a few companies leave the ETF, weight plummeting to zero, and a few come in, with the same weight as the largest companies in the ETF. Apart from the irrationality, that is a lot of trading costs to deal with.
- Similarly, at every rebalance point stocks have to be traded to yank them back to weight. Again, very high cost compared to cap weighted rebalancing.
- Equal weight cannot scale and as the fund gets larger there is increased risk of other market participants front running required purchase/sales ahead of a rebalance.

[Newroad]

Hi HiRiskPaul.

Let's assume both a hypothetic S&P500 Capital Weight ETF (CAPETF) and Equal Weight ETF (EQUETF).

I believe it's evident by inspection that they will produce different outcomes, therefore they both can't be market neutral. So, if you assert that CAPETF would be market neutral, EQUETF cannot be. This would presumably make EQUETF some form of active investing, which is in effect what GeoffF100 asserts. Conversely, as you know, I believe both are forms of passive investing, with the nature of the market neutrality reflecting the specific type of passive investing.

I judge your perception of CAPETF being market neutral is based more on precedence/commonality, rather than theoretic merit - but I'm of course happy to be educated to the contrary

Regards, Newroad

PS I note your subsequent post and most points raised are fair. Nothing is perfect - you pay your money and take the (imperfect) ride. I noted early on in this discussion equal weighting may either be not worth doing or too hard to do.

[GeoffF100]

Tracker funds already hold more than half the US market.

Equal weighting over weights the smaller cap stocks. If >50% of the market bought these funds the price of the smaller weighted stocks would be huge. It would be crazy to buy such a fund in those circumstances. You may think that is not a significant matter in the current market, and perhaps you are right. Nonetheless, you would be making a market judgement and that is passive investing.

Similarly for low volatility funds. If they are over-invested they become a bad choice. You need to make a judgement as to whether they are over invested before investing in them, and that is again active investing.

The only purely passive approach is to buy a fixed percentage of everything - i.e a market weighted tracker.

[hiriskpaul]

Hi HiRiskPaul.

Let's assume both a hypothetic S&P500 Capital Weight ETF (CAPETF) and Equal Weight ETF (EQUETF).

I believe it's evident by inspection that they will produce different outcomes, therefore they both can't be market neutral. So, if you assert that CAPETF would be market neutral, EQUETF cannot be. This would presumably make EQUETF some form of active investing, which is in effect what GeoffF100 asserts. Conversely, as you know, I believe both are forms of passive investing, with the nature of the market neutrality reflecting the specific type of passive investing.

I judge your perception of CAPETF being market neutral is based more on precedence/commonality, rather than theoretic merit - but I'm of course happy to be educated to the contrary

Regards, Newroad

PS I note your subsequent post and most points raised are fair. Nothing is perfect - you pay your money and take the (imperfect) ride. I noted early on in this discussion equal weighting may either be not worth doing or too hard to do.

The terms "Passive" and "Active" are labels we are stuck with, but they are not very useful descriptions of the types of investing they intend to categorise. I prefer Systematic and Unsystematic. Systematic investing means buying and selling according to a prescribed set of rules. Unsystematic investing brings in human judgement, or maybe AI one day.

Cap weighting is special form of Systematic investing and is reasonably deserving of the term "Passive". To see why it is easier to get away from the "Cap weighted" label initially. The S&P 500 is not actually a true cap weighted index by the way as its constituents are chosen by a committee, but it is close enough to behave like a cap weighted fund of the largest 500 companies.

If you form a portfolio by buying 1% of the shares of every company in the market, then I think you would agree that this reflects the market as a whole. Your portfolio would precisely follow the total value of the market up and down as stock prices fluctuated. Your portfolio would slowly go out of line as companies joined the market, existing companies issued new shares and others bought them back, so you would need to regularly update the portfolio to ensure you had exactly the same proportion of the shares in each company. There are other minor details that need to be handled, such as the situation in which not all shares are freely available for purchase on the market, but for now just assume you buy hold and adjust so that you hold the same proportion of shares in each company.

This portfolio would be truly reflective of the market as 1% of the shares of each company would rise and fall exactly in line with the market as a whole. This portfolio is equal weight by proportion of shares. It is absolutely not equal weight by value as some companies are worth substantially more than others, and prices change all the time. As well as being market neutral, this "equal weight by proportion of shares" portfolio is fairly easy to keep in line as the total number of shares in each company does not fluctuate that much. Guess what? This portfolio is the cap weighted portfolio! The portfolio that best reflects the whole market is the whole of market cap weighted portfolio and is what makes cap weighting special compared with other systematically managed portfolios.

[1nvest]

The only purely passive approach is to buy a fixed percentage of everything - i.e a market weighted tracker.

You can equally buy a fixed capital amount of everything, or a fixed number of shares of everything and remain passive. In practice there is no passive choice, as stocks will flow into/out of the set/index. In some cases a index may be changed, such as when the S&P500 was transitioned over to being 'domestic only' that had index funds having to sell some shares to buy others. Fixed % isn't consistently > fixed capital or fixed shares, longer term indications are that fixed capital > fixed %. Some suggest that's a size factor issue. Personally I suspect its that fixed % tilts towards certain stocks and as most stocks under-perform the broader average due to the average being uplifted by the few that do incredibly well, that you broadly lose out due to not having weighted each equally capital wise. If I capital weight A twice as heavily as B, then great if A is the winner, but otherwise I had less invested in the winning B stock than had I weighted both equal capital amounts.

Fixed % does help market products to the broad masses, but as individual investors its not the best. Spreading risk/rewards equally across the individuals from the offset is the more neutral stance rather than buying more into what historically was the better performers. Look at those that were tech heavy prior to the dot com bubble bursting, or financials heavy just prior to the 2008/9 financial crisis. Even the recent S&P500 that reflects 'US' and that is around half of the world cap weighted has 5 tech stocks making up 18% of the index value. Going forward another sector might be the relative winner, perhaps pharmaceuticals for instance. Placing what is paramount to a bet that techs will at least rival or better pharmaceuticals as that makes funds marketing easier is in their best interest, not yours.

[1nvest]

How do you measure the 'market'. Most/many use the 'Index'. But the market index has changed/evolved over time. Dow and Jones devised three indexes, Transport, Utilities and Industrial averages, but that evolved into a single choice, the best outcome choice. Investors who tracked all three initial indexes would have lost out due to that evolution. Indexes are a mathematical model that have evolved over time, where the recent best choice is then often reflected back as a suggested indicator of past results, but where that mathematical model isn't known in advance. Only by predicting or by luck implementing that model can you achieve that 'average' in practice (excluding costs).

In the context of all stocks being reflective of bakery shops then yes, it seems reasonable to weight one stock that owned 30 shops thirty times more than another that reflected just one shop. The single shop stock however has greater/easier potential to double its size. More broadly it is not all just a single sector/business, and for some stocks their market cap can reflect massive amount of intangible assets, little real/tangible assets value where that value could just vapourise overnight. There are potential future 'market average' mathematical models that might come to replace existing preferred models as the broad measure of the 'market average'.

[hiriskpaul]

Tracker funds already hold more than half the US market.

Equal weighting over weights the smaller cap stocks. If >50% of the market bought these funds the price of the smaller weighted stocks would be huge. It would be crazy to buy such a fund in those circumstances. You may think that is not a significant matter in the current market, and perhaps you are right. Nonetheless, you would be making a market judgement and that is passive investing.

Similarly for low volatility funds. If they are over-invested they become a bad choice. You need to make a judgement as to whether they are over invested before investing in them, and that is again active investing.

The only purely passive approach is to buy a fixed percentage of everything - i.e a market weighted tracker.

Minimum volatility is a market anomaly that goes back a long way. It is not easy to exploit as the volatility of stocks needs to be constantly monitored, selected/deselected. As well as sorting by volatility, selected stocks need to show low correlation with each other, so as to further reduce portfolio volatility, all whilst keeping an eye on diversification and not introducing too much portfolio churn. Unlike small caps or value, there is no clear cut way of exploiting the anomaly and different fund managers will follow different strategies. However, if low/minimum vol becomes too popular then there is of course a risk that the anomaly will disappear. I really should stop talking about it

I would definitely not recommend low/min vol as a farm bet, but I am happy to have invested in USMV. I nearly chose a value fund instead, which would not have worked out well. Provided I continue to get lower than market risk adjusted returns, I would be content to lag the market, as USMV has at times. Part of the motivation for buying USMV was the old DC capped income drawdown regime. Irrelevent now, but I still appreciate the lower volatility.

[hiriskpaul]

How do you measure the 'market'.

That's easy. Multiply the number of shares of each company in the market by the share price. That gives the instantaneous value of each company. Then add up all those company values.

Cap weighting is a construction that attempts to closely follow this market valuation. Hard to see how it could be improved upon.

[GeoffF100]

The "market portfolio" is the same percentage of every investment available. That is the only portfolio that does not distort market prices if it is bought in enormous quantity. You would have bought an awful lot of tulips in the tulip mania, which would not have been good. Nonetheless, if you had been steadily investing over a long period of time, that would not have been a problem. Large cap stocks can be over-valued just a small cap stocks can, but there is no reliable way of knowing when to get out. I am not saying that buying the market portfolio (or even its equity part) will have the best outcome, but it is the only choice that does not require any judgement about market values.

[hiriskpaul]

Fixed % does help market products to the broad masses, but as individual investors its not the best. Spreading risk/rewards equally across the individuals from the offset is the more neutral stance rather than buying more into what historically was the better performers. Look at those that were tech heavy prior to the dot com bubble bursting, or financials heavy just prior to the 2008/9 financial crisis. Even the recent S&P500 that reflects 'US' and that is around half of the world cap weighted has 5 tech stocks making up 18% of the index value. Going forward another sector might be the relative winner, perhaps pharmaceuticals for instance. Placing what is paramount to a bet that techs will at least rival or better pharmaceuticals as that makes funds marketing easier is in their best interest, not yours.

What would have happened in the dot-com bubble if you had equal weighted the myriad dot-com companies that subsequently turned out to be worthless?

[Newroad]

That's an interesting question, HiRiskPaul.

Wouldn't one have been, in effect, top slicing them on the way up?

Would be an interesting back-test - might well have been an OK outcome.

Regards, Newroad

[Lootman]

RSP chooses to equal weight the largest 500 US companies. But if we are equal weighting, why choose those? What about the next 500 companies which have achieved zero weighting according to this scheme?

You hit a practical problem if you try and equal weight a much broader index e.g. the Russell 3000.

And that is because the fund cannot buy the tiny shares in the same amounts as it can the larger shares. The tiny shares just do not have the liquidity and breadth to successfully equal-weight them. Even trying to do that would distort their prices.

So in practice equal-weighting only works with larger and more liquid securities. And as I mentioned earlier perhaps its best use is with sector funds where one or two shares unduly dominate that sector.

[hiriskpaul]

That's an interesting question, HiRiskPaul.

Wouldn't one have been, in effect, top slicing them on the way up?

Would be an interesting back-test - might well have been an OK outcome.

Regards, Newroad

I certainly don't know the answer. Equal weight will reduce the allocation to a few large companies but will also increase the allocation to many more small ones. Who knows whether that would have made the fallout better or worse.

In moving away from whole of market cap weighting, you really have to ask yourself what you hope to achieve and look for evidence to back up whether your deviation is at all rational or justified. One frequently used deviation is going for actively managed funds. There is overwhelming evidence now that doing that will reduce your expected risk adjusted returns. That is not to say you will not have a good outcome by pure dumb luck, but that is all you can hope for with actively managed funds. At any point in time there will always be some funds showing outstanding risk adjusted returns (Buffett, etc.) and it is very tempting to assume the performance of these funds/managers will continue, although there is copious evidence that this is unlikely. Woodford being a recent example of how decades long good performance can evaporate.

There is reasonable evidence that some market anomalies do exist and are exploitable. Size, value, quality, momentum, low vol, etc. but I have never come across a "Equal Weight Factor" in any of the academic literature, nor can I see any logical reason why it might give better risk adjusted returns than a whole of market tracker, except perhaps by overweighting size and value when size and value do well. Equal weight is going to set sector risk according to the number of companies in a sector. I cannot see the logic in that and it is totally unclear to me why that might be a good thing to do. A clear disadvantage is the cost of maintaining the weighting.

Another problem with moving away from cap weight, evidence backed or not, is that market out performance is not guaranteed and even with the various evidence based anomalies/factors, they don't work all the time. Value investing has been dire for a decade. For anyone sitting outside cap weighting and seeing a few years of under performance, there is then the dilemma of whether to hold the course or to jump back into cap weighting (or on to something else showing good recent performance).

If you want to increase the weight of smaller caps, for a reasonable prospect of higher returns (at higher risk), that is easy to do via a small cap fund. It is harder though to reduce the weight of the mega-caps and I can see the potential desirability in doing that in some situations. Nokia dominates the Helsinki exchange and is capped at 10% in the OMX Helsinki 25 for this reason. If any one company rose to over 10% of the World index I think I would get concerned about all-in cap weighting, unless a cap was introduced.