Author Topic: Active Super Management beats Passive Super Mgmt Over the Long Term? (Australia)  (Read 4007 times)

This_Is_My_Username

  • Bristles
  • ***
  • Posts: 323
  • Location: Australia, Mate.
The November 2014 issue of Money Magazine had this article (transcribed):

Strong Skills are Worth Paying For

Should Your Super be Handed to Active or Passive Managers?

Australian Super funds typically are more than 50% invested in shares, signifacntly higher than most countries' retirement fund schemes.  Shares can be more expensive to manage and there is much debate over if it is worth paying more for active rather than passive management.  Super funds are very different from the retail funds management industry because they are wholesale investors with scale and negotiating power.  This means the fees they pay for active management are much lower than the fees paid by retail investors.  For example, they pay about 0.3% and 0.5% for actively managed australian shares and international shares, respectively, which is about the same a retail managed funds investors pay for passive management.  As a result, super funds have a greater chance of generating excess returns after fees. 

While not all active managers outperform the index, our research shows there are enough managers with the necessary skills to generate excess returns (net of wholesale fees). While there is clearly a place for passive management, mandating wholly passive management would be a mistake and would take away an important tool that funds have used to add value for members. 

The investment performance of the 10 largest MySuper products - with a total of $360 billion under management and 10 million members - has been strong.  As a group, they have outperformed the passive benchmark bu 0.4%pa, 0.6%pa, and 0.8%pa over the seven, 10 and 15 years to June 2014, respectively.  To put this in perspective, the 0.8%pa outperformance over 15 years translate to a cumulative outperformance of about 20%.  That makes for a substantial diference in a member's nest egg. 

This outperformance stems from two main sources: asset allocation and stock selection.  These 10 super funds typically have 15% invested passively, the bulk of it in Australian (7%) and international shares (4%).  The typical premium for actively managed australian shares is about 0.3% pa.  Large super funds can negotiate even lower fees.  Clearly, the premium for active management has been justified, suggesting that there are enough managers with the appropriate skills to have a reasonable likelihood of adding value. 

Warren Chant
(founded super research house ChantWest in 1997, primarily to advise medium to large employers on outsourcin their funds)

This_Is_My_Username

  • Bristles
  • ***
  • Posts: 323
  • Location: Australia, Mate.
.
« Reply #1 on: April 10, 2015, 06:38:38 AM »

So I've recently read "A Random Walk Down Wall Street".  Which persuaded me that active management can't outperform passive management over the long term.  And most members of this forum would agree. 

But this article has a different view.

Any insights in to this discrepancy?

Cheers

Aphalite

  • Bristles
  • ***
  • Posts: 425
Let's get this out of the way first - Passive = Allocate capital based on float or market cap % of total market of the company, Active = Allocation of capital in any other way, equal weight is one example, holding only 20 companies that the manager believes is better than the benchmark is another example

Active can't beat passive most of the time because of costs/fees - 1) cost of turnover, when active management sells stocks, it triggers capital gains and these are taxed 2) management/advertising fees (the ones often cited in these forums, usually 1-2%) - if these Australian active funds charge around the same fees as passive Australian funds, I can see how active could beat passive

The oft cited study of how active management doesn't beat passive management actually finds that 80% of active managed funds MEET the passive benchmarks before fees, but only 10-20% beat it after accounting for fees - this begs the question (at least in the US) - why pay for active management at all?

If Australia's structure has active funds and passive funds costing the same, active is theoretically better (as far as picking the right one, good luck)

qwerty8675309

  • 5 O'Clock Shadow
  • *
  • Posts: 51
I can't see how the MySuper products can beat the passive benchmark over 15 years. The MySuper products have only been around since 2011 when the Gillard government introduced them.

Without a lot of the detail around how they've worked out these performance figures, and the type of benchmark they are using, it's difficult to critique the article. What I think is important to keep in mind is that the market is a zero sum game (or a negative sum game after fees). A winning trade is offset by a losing trade. If the top 10 biggest super funds are beating a standard passive benchmark (which I highly doubt), that must mean an equal number of trades are losing by the same amount.

Aphalite

  • Bristles
  • ***
  • Posts: 425
What I think is important to keep in mind is that the market is a zero sum game (or a negative sum game after fees). A winning trade is offset by a losing trade. If the top 10 biggest super funds are beating a standard passive benchmark (which I highly doubt), that must mean an equal number of trades are losing by the same amount.

No no no. We need to stamp out that notion around here. If the market is a zero sum game, it would not always go up in the long run. Companies generate earnings which are then reinvested to make the company more valuable or distributed to owners (sometimes via buy backs which increases stock price without harming any other owners). The only reason people compare to the benchmarks is because it's the current popular baseline. At times of high inflation, a better opportunity cost to compare to could be treasury bonds, which used to yield double digits in the 80s. But benchmarks or index funds themselves have pre established HUMAN criteria, it's not an entity, just a collection of companies by float. Some companies with higher float could be less valuable or generate less revenues and income than companies with less float

Indexer

  • Handlebar Stache
  • *****
  • Posts: 1437
What I think is important to keep in mind is that the market is a zero sum game (or a negative sum game after fees). A winning trade is offset by a losing trade. If the top 10 biggest super funds are beating a standard passive benchmark (which I highly doubt), that must mean an equal number of trades are losing by the same amount.

No no no. We need to stamp out that notion around here. If the market is a zero sum game, it would not always go up in the long run. Companies generate earnings which are then reinvested to make the company more valuable or distributed to owners (sometimes via buy backs which increases stock price without harming any other owners). The only reason people compare to the benchmarks is because it's the current popular baseline. At times of high inflation, a better opportunity cost to compare to could be treasury bonds, which used to yield double digits in the 80s. But benchmarks or index funds themselves have pre established HUMAN criteria, it's not an entity, just a collection of companies by float. Some companies with higher float could be less valuable or generate less revenues and income than companies with less float

Let me phrase what qwerty was trying to say a little better.  The market isn't a zero sum game.  Trying to beat the market is.

All the investors in the market = the market.  For one investor to outperform one investor must underperform.  So if the market is up 10% and one guy is up 11% someone out there is only up 9%.  Now throw in 1% fees, 11 turns to 10, 9 turns to 8.  As a group active investors will underperform the market because of fees and taxes.  It is simple math.


Quote from: This_Is_My_Username
This outperformance stems from two main sources: asset allocation and stock selection.  These 10 super funds typically have 15% invested passively, the bulk of it in Australian (7%) and international shares (4%).  The typical premium for actively managed australian shares is about 0.3% pa.  Large super funds can negotiate even lower fees.  Clearly, the premium for active management has been justified, suggesting that there are enough managers with the appropriate skills to have a reasonable likelihood of adding value. 

Asset allocation is not a way for a 'fund' to the beat the market.  You can't achieve Alpha by changing your asset allocation because your benchmark would also change with the asset allocation.  American funds pulls this crap.  A fund that is 90% domestic equity and 10% international beats the 500 index in a year when international does really well and suddenly they are beating the market.... no they aren't.  The 90% domestic underperformed the 500 index, and the 10% international outperformed the 500 index but it still underperformed the international market.  You can 'say' you beat the benchmark but it is because its the wrong benchmark and your Alpha was still negative.  Sorry for the rant... this is one of my pet peeves.  If one of their sources of outperformance is 'asset allocation' then they aren't really outperforming... they just have the wrong benchmark.
« Last Edit: April 10, 2015, 08:47:56 PM by Indexer »

Aphalite

  • Bristles
  • ***
  • Posts: 425
Indexer, I completely agree that in the US, active loses due to taxes and fees, so why pay for it? Based on the op's excerpt, it seems that fees are comparable to a passive fund, unlike the US

On the other hand, I still disagree that "beating" the benchmark is zero sum. The benchmark is a very specific set of guidelines that could give you long term alpha versus an equal weighted index or could under perform an equal weighted index. As you said, everyone has a different benchmark and different risk portfolio/toleration, it is entirely possible for more or less than 50% of market participants to beat their benchmarks. It is quite improbable if you are paying fees or being charged taxes due to turnover tho
« Last Edit: April 10, 2015, 09:02:03 PM by aphalite »

This_Is_My_Username

  • Bristles
  • ***
  • Posts: 323
  • Location: Australia, Mate.
I can't see how the MySuper products can beat the passive benchmark over 15 years. The MySuper products have only been around since 2011 when the Gillard government introduced them.

Without a lot of the detail around how they've worked out these performance figures, and the type of benchmark they are using, it's difficult to critique the article. What I think is important to keep in mind is that the market is a zero sum game (or a negative sum game after fees). A winning trade is offset by a losing trade. If the top 10 biggest super funds are beating a standard passive benchmark (which I highly doubt), that must mean an equal number of trades are losing by the same amount.

Many MySuper products were created by simply re-naming an existing 'balanced' investment option, because that option already met the MySuper criteria.    That is how they can have decades of historical results. 

Indexer

  • Handlebar Stache
  • *****
  • Posts: 1437
Indexer, I completely agree that in the US, active loses due to taxes and fees, so why pay for it? Based on the op's excerpt, it seems that fees are comparable to a passive fund, unlike the US

On the other hand, I still disagree that "beating" the benchmark is zero sum. The benchmark is a very specific set of guidelines that could give you long term alpha versus an equal weighted index or could under perform an equal weighted index. As you said, everyone has a different benchmark and different risk portfolio/toleration, it is entirely possible for more or less than 50% of market participants to beat their benchmarks. It is quite improbable if you are paying fees or being charged taxes due to turnover tho

I think we have different definitions of the word benchmark.  Your benchmark should represent what you are investing in.  If you are 10% small caps, 50% large caps, 30% international, 20% bonds... you should use a benchmark that is made up of indexes representing those investments in those allocations. Beating the 'market' is zero sum.  Its math.  Investors as a group can't average returns higher than the collective of the things they are invested in.  If the returns of the market are 100 billion, investors can't get 101 billion. 

You also can't achieve Alpha by investing in something other than your appropriate benchmark... whatever that is.  You can only measure Alpha by first having an appropriate benchmark.  alpha=portfolio return-[risk free rate+(market return-risk free rate)beta].  If you don't have the appropriate benchmark to use as your 'market return' then your R^2 will be to low to use beta.  Without beta you can't calculate alpha.  This brings me to my previous point; you can't achieve alpha through 'asset allocation.'  Changing your asset allocation changes what you should be comparing yourself against.  The 'market return' and beta in the alpha equation will both change to represent the new asset allocation(market you are comparing against).  Otherwise you are just making up numbers and it makes zero sense.

And 0.3-0.5% isn't 'low' to me.  Vanguard passive funds are normally under 0.1% for domestic and under 0.2% for international.  Vanguard also has active funds in the 0.2-0.5% range.  They also tend to perform very well.

This_Is_My_Username

  • Bristles
  • ***
  • Posts: 323
  • Location: Australia, Mate.
.
« Reply #9 on: April 11, 2015, 08:58:56 PM »
Quote
And 0.3-0.5% isn't 'low' to me.

things are different in Australia, etf fees are higher.

vanguard are still the cheapest.

http://en.wikipedia.org/wiki/List_of_Australian_exchange-traded_funds

https://www.vanguardinvestments.com.au/retail/ret/investments/etfs.jsp#etfstab

Aphalite

  • Bristles
  • ***
  • Posts: 425
Beating the 'market' is zero sum.  Its math.  Investors as a group can't average returns higher than the collective of the things they are invested in.  If the returns of the market are 100 billion, investors can't get 101 billion. 

You also can't achieve Alpha by investing in something other than your appropriate benchmark... whatever that is.  You can only measure Alpha by first having an appropriate benchmark.  alpha=portfolio return-[risk free rate+(market return-risk free rate)beta].  If you don't have the appropriate benchmark to use as your 'market return' then your R^2 will be to low to use beta.  Without beta you can't calculate alpha.  This brings me to my previous point; you can't achieve alpha through 'asset allocation.'  Changing your asset allocation changes what you should be comparing yourself against.  The 'market return' and beta in the alpha equation will both change to represent the new asset allocation(market you are comparing against).  Otherwise you are just making up numbers and it makes zero sense.

And 0.3-0.5% isn't 'low' to me.  Vanguard passive funds are normally under 0.1% for domestic and under 0.2% for international.  Vanguard also has active funds in the 0.2-0.5% range.  They also tend to perform very well.

First part is only true if you are assuming no new money is being added to the "market". Consider the simplest scenario possible. If an investor is holding $100 worth of stock and that was all that existed in the market, and sold for $120 at the end of the year, he got a 20% return for the year and the benchmark would be 20% because there was only one stock. The investor he sold it to bought it at the end of the year at 120, and that's where the price ended as there were no other transactions. His return is not 0% for the year, because he didn't have any money invested during the year, not because he was holding cash, but because he didn't get paid until 12/31. This is a simple scenario but I hope it illustrates my point - as I said, if the "market" was a zero sum game, stocks wouldn't always go up - the businesses that make up the "market" generates value, which makes everyone richer, allowing them to further invest their acquired wealth

When you discuss alpha and beta, you have to make the underlying assumption that everyone in the market is operating with the same information and with the same investing skill set, and that stock prices are representative of the true value of the underlying enterprises. This is not true for the short or even intermediate time frames. Beta or price volatility, is only pertinent if you are trading short term options or on short term price speculation. It does not present RISK in the sense that your purchasing power is reduced as compared to opportunity cost of investing your capital elsewhere. Beta is a nice academic number that people like to use because it's elegant and fits theoretically, much like a "zero sum" theory, but if you are choosing between a stock that returns 3% in dividends and never change in price, vs a stock that gyrates up and down but eventually ends up appreciating 20% in price, beta will say A is the best and B is a terribly risky investment, without giving any thought to capital structure of the companies, customer diversification, etc. Or put another way, a good measure of risk should tell you "what is the possibility that my purchase power will decrease in comparison to other investments and inflation?". Beta does not do this because it's qualitatively blind - in fact, if a stock gets cheaper due to panic and you can buy more of it at a lower price (in the process building in a margin of safety), it tells you that somehow the stock has become MORE risky

.3% might be high in your estimation, but from the OP's post I assumed that people in Australia do not have access to vanguard funds, and so they must choose between the opportunity costs of what they CAN invest in (article said passive etfs in Australia are .3-.5%). If I am wrong, then the point is moot, they are probably better off with a passive index. My point is if passive funds and active funds cost the same, then you have lost one of the biggest advantages that passive indices have over active management, the other being lack of turnover
« Last Edit: April 11, 2015, 10:15:07 PM by aphalite »

Indexer

  • Handlebar Stache
  • *****
  • Posts: 1437
Beating the 'market' is zero sum.  Its math.  Investors as a group can't average returns higher than the collective of the things they are invested in.  If the returns of the market are 100 billion, investors can't get 101 billion. 

You also can't achieve Alpha by investing in something other than your appropriate benchmark... whatever that is.  You can only measure Alpha by first having an appropriate benchmark.  alpha=portfolio return-[risk free rate+(market return-risk free rate)beta].  If you don't have the appropriate benchmark to use as your 'market return' then your R^2 will be to low to use beta.  Without beta you can't calculate alpha.  This brings me to my previous point; you can't achieve alpha through 'asset allocation.'  Changing your asset allocation changes what you should be comparing yourself against.  The 'market return' and beta in the alpha equation will both change to represent the new asset allocation(market you are comparing against).  Otherwise you are just making up numbers and it makes zero sense.

And 0.3-0.5% isn't 'low' to me.  Vanguard passive funds are normally under 0.1% for domestic and under 0.2% for international.  Vanguard also has active funds in the 0.2-0.5% range.  They also tend to perform very well.

First part is only true if you are assuming no new money is being added to the "market". Consider the simplest scenario possible. If an investor is holding $100 worth of stock and that was all that existed in the market, and sold for $120 at the end of the year, he got a 20% return for the year and the benchmark would be 20% because there was only one stock. The investor he sold it to bought it at the end of the year at 120, and that's where the price ended as there were no other transactions. His return is not 0% for the year, because he didn't have any money invested during the year, not because he was holding cash, but because he didn't get paid until 12/31. This is a simple scenario but I hope it illustrates my point - as I said, if the "market" was a zero sum game, stocks wouldn't always go up - the businesses that make up the "market" generates value, which makes everyone richer, allowing them to further invest their acquired wealth

alphite.  Here is the confusion.  I'm not saying the stock market is a zero sum game.  I'm saying BEATING the market is a zero sum game, and I've been pretty specific about that.  The market goes up over time, thats the point.  I am NOT saying for each investor who earns 10% someone earned -10%.  I'm saying if the market earned 10% and 1 investor earned 11% someone else earned 9%.  Beating the market is a zero sum game.  For each dollar that 'won' and beat the market another dollar 'lost' and underperformed the market.  If the market is up 10% the 'average' investor earned 10% before fees/taxes.  Bogle founded Vanguard based on that idea and he has written several books about it.

Quote
When you discuss alpha and beta, you have to make the underlying assumption that everyone in the market is operating with the same information and with the same investing skill set, and that stock prices are representative of the true value of the underlying enterprises. This is not true for the short or even intermediate time frames. Beta or price volatility, is only pertinent if you are trading short term options or on short term price speculation. It does not present RISK in the sense that your purchasing power is reduced as compared to opportunity cost of investing your capital elsewhere. Beta is a nice academic number that people like to use because it's elegant and fits theoretically, much like a "zero sum" theory, but if you are choosing between a stock that returns 3% in dividends and never change in price, vs a stock that gyrates up and down but eventually ends up appreciating 20% in price, beta will say A is the best and B is a terribly risky investment, without giving any thought to capital structure of the companies, customer diversification, etc. Or put another way, a good measure of risk should tell you "what is the possibility that my purchase power will decrease in comparison to other investments and inflation?". Beta does not do this because it's qualitatively blind - in fact, if a stock gets cheaper due to panic and you can buy more of it at a lower price (in the process building in a margin of safety), it tells you that somehow the stock has become MORE risky

Alphite, you brought Alpha into this, not me ;).  You can't calculate Alpha without Beta so that is where that came from.  Let me point out something vital.  When you are calculating alpha you are doing it AFTER the fact.  It is not an equation used to predict the future so I'm not sure what you were getting at with this paragraph.  It looks like a long equation but really it is just Alpha = actual portfolio return - what a portfolio with similar risk 'should' have returned.  Positive alpha means you did better than you should have, negative alpha means you did worse than you should have done.  When you are looking at data after the fact its not really theory or predictions anymore.  You are looking at cold hard facts. 

superannuationfreak

  • Stubble
  • **
  • Posts: 126
  • Location: Melbourne, Australia
There are a lot of nuances in this issue that don't easily fit in a 350 word article targeted at a general audience.  There are a few different issues which are muddled together here, so let's see if we can disentangle them a bit.

On asset allocation, most of the large MySuper funds being discussed are from industry funds.  Most of them have a significant allocation to direct infrastructure and property, neither of which have been that well-proxied with passive approaches historically - it is reasonable to ask whether it will continue in the future with more competition for these assets, or whether the lack of liquidity in these assets means we have not been paid a sufficient premium to hold them, but in the one path of history we have experienced it has been a successful strategy.

Then we get into specific asset classes.  We should be cautious here too.  On the one hand for Australian Shares and Bonds, industry funds have been able to keep costs low and pick good investment management.  Based on more recent performance I'd argue it is getting harder to outperform in Australian Shares, but when you can for instance get Australian Super's Australian Share investment option for 0.30% it is unlikely to be a big mistake (but equally I would not expect large outperformance given the size of their Australian Share portfolio, nor would I usually pay much more than that).  On Bonds, there is only one widely cited passive benchmark used in Australia (from UBS, plus S&P use their own) so extending maturity and/or credit risks can often be rewarded at some points in time.  Broader bond indices (including non-financial corporate bonds) aren't as easily replicated as share indices, and even Vanguard in the US has many of their bond funds as non-index funds (but still focused on keeping costs low, much like the better industry funds).

Other asset classes have a less impressive record which is diluted by using figures for MySuper/balanced funds.  In particular, I'm not convinced that Australian superannuation funds have done such a good job at International Shares and passive costs are now around 0.2% p.a. or much less (although no super fund I know of has a non-direct offering less than about 0.3% p.a. including admin costs).  To pay more than that I'd want more transparency around what we're targeting (for example, if targeting the Value factor I'd want to know what we're getting that DFA couldn't provide for 0.3-0.4% p.a.).  Similarly, I'm unconvinced that Australian superannuation funds have premium access to the best private equity or hedge funds (in infrastructure we seen to have local expertise) and without access to/ability to find the best of these we know the average contribution provided by these funds to a portfolio is unlikely to be a great one.

Also, keep in mind that most for-profit funds still charge more on a like-for-like basis.  For example when you compare Australian Super's Balanced fund (around 0.6%p.a. fees) to ANZ's Smart Choice Super (0.5% p.a.) you're ignoring that 20-25% of the Australian Super fund is direct property and infrastructure with whatever costs and benefits that comes with.  A fairer comparison would be with Australian Super's Indexed Diversified fund at 0.21% p.a.

So, my TL;DR: There is likely some alpha in Australian Share portfolios held by these funds but I expect that to continue to decline over time.  There is more competition for good infrastructure assets now but there is still probably an expertise 'edge' that the industry funds have here, so if you believe access to direct infrastructure and property is valuable then industry funds seem a good place to get that exposure.  But I'm unconvinced on assets for which there is much greater global competition (international shares, private equity and hedge funds)

This_Is_My_Username

  • Bristles
  • ***
  • Posts: 323
  • Location: Australia, Mate.
.
« Reply #13 on: April 25, 2015, 05:58:44 AM »
cheers mate, thanks for providing your valuable input : )