Porfolio composition discussion

[Lagrangian Mechanic]

In light of David's recent post (I am not the poster he was replying
to) I figured I'd put up part of my asset class breakdown for discussion,
ripping apart, etc.

My personal methodology is to first split things into fixed-income
and non-FI. The non-FI piece is split into domestic non-FI and
foreign non-FI. When I do change my risk assessment (which is not
often), it is changed by dialing those to knobs. In other words,
the domestic non-FI piece always has the same composition, but the
amount of the domestic non-FI piece in the total portfolio can
change.

The discussion I'm looking for is limited to the components
of the domestic and foreign non-FI pieces.

My personal ground rules:
* Investments have to be directly available to retail
investors (so, sadly, no DFA funds for example)
* Passive index funds except when none exist in the
asset class
* No ET*N*s. I'm not looking to take on the
issuer's credit risk.
* Nothing that results in me getting K-1s or UBTI

Primary things I'm interesting in hearing about:
* Better proxies for the asset classes in question
than the ones I'm using.
* Missing asset classes (and proxies for them) to
consider.
* Changes in allocation between asset classes

Domestic non-FI component (listed %ages are %ages
of this component, not of all non-FI and not of total
portfolio):

Small-Cap (IWM) - 20%
Small-Cap Value (IWN) - 20%
Large-Cap (IWB) - 25%
Large-Cap Value (IWD) - 25%
Micro-Cap (IWC) - 10%
Micro-Cap Value (???) - 0%

Foreign non-FI component (listed %ages are %ages
of this component, not of all non-FI and not of total
portfolio):

Int'l Large-Cap (EFA) - 25%
Int'l Large-Cap Value (EFV) - 25%
Int'l Small-Cap (GWX) - 20%
Int'l Small-Cap Value (DLS) - 20%
Emerging Markets (EEM) - 10%
Emerging Markets Value (???) - 0%

While I won't blame the pros here for not giving
detailed into for free, if they feel inspired
to do so, I certainly won't complain! :)

--
Lagrangian

[dumbstruck]

I will shut up about my personal view, but point out some probably valuable resources to compare and contrast. Here is a clear table of the current and target asset allocations of CalPERS pension fund http://www.calpers.ca.gov/index.jsp?bc=/investments/assets/assetallocation.xml

I earlier posted a similar one by Northern Trust Bank and probably many institutions publish them (university endowments, sovereign funds?). Of course they don't fully live by your preconditions or goals, but they are recognized centers of expertise. Oh here is CalSTRS which is swinging for the fences in Japanese equity: http://www.calstrs.com/current-investment-portfolio

CNBCny today interviewed CIO's of CalPERS and the rival CalSTR together today and I believe it was CalSTRS upping the risk to overcome low funding and falling behind CalPERS. They take comfort that the taxpayer makes up any money they lose, so maybe not the best example to follow.

[Tad Borek]

On 4/30/2013 4:31 PM, Lagrangian Mechanic wrote:
> Primary things I'm interesting in hearing about:
> * Better proxies for the asset classes in question
> than the ones I'm using.
> * Missing asset classes (and proxies for them) to
> consider.
> * Changes in allocation between asset classes

Asset classes...Allocation to "non-large" stocks is large - do you
believe very strongly that smaller-cap stocks, in general, will perform
better over the long haul (including in foreign markets)? Your mix
suggests that you do, which is fine, but if you don't - why so high? And
did you intentionally omit midcaps? You'd get them with a "total market"
kind of core holding.

Investments...nothing specific, just generally...using that many ETFs
could lead to a lot of trading costs if you plan to round-trip buy/sell
to rebalance the mix, or invest or withdraw smaller amounts
periodically. Owning traditional no-load mutual funds through the fund
company can reduce that cost. Not just commissions, but also trading
spreads and discount/premium to NAV. Also, when looking at funds vs.
ETFs I have concerns about how ETFs will perform in less-liquid asset
classes, especially when things get interesting. I don't know how the
whole ETF creation/redemption process could hold up in a period of high
volatility or inflows/outflows, if the ETF holds stuff that is hard to
trade (and by extension hard to value every millisecond).

-Tad

[David S Meyers CFP]

On 2013-05-01 18:39:35 +0000, Tad Borek said:

> On 4/30/2013 4:31 PM, Lagrangian Mechanic wrote:
>> Primary things I'm interesting in hearing about:
>> * Better proxies for the asset classes in question
>> than the ones I'm using.
>> * Missing asset classes (and proxies for them) to
>> consider.
>> * Changes in allocation between asset classes
>
> Asset classes...Allocation to "non-large" stocks is large - do you
> believe very strongly that smaller-cap stocks, in general, will perform
> better over the long haul (including in foreign markets)? Your mix
> suggests that you do, which is fine, but if you don't - why so high?
> And did you intentionally omit midcaps? You'd get them with a "total
> market" kind of core holding.

He's got 50% large (split between value and blend) and 50% small/micro.

There's some vagueness - mid-caps fall where? - but it's not an
outrageous oveweight to small-caps.

By comparison, and using Morningstar size breakdowns, Flexshares TILT
-- which is intented to emphasize small-cap and value (ie. Fama/French
factors) -- is about the same breakdown: 50% large (including mega) and
50% small (including most of mid-caps).

The total market, with market-weight indices (ie. Vanguard's VTI) is
about 71% large/mega and 29% micro/small/mid.

That's a big difference, but it's not huge. I see folks come in all
the time with vast overweights one way or the other (though more often,
folks are overweight large/mega and only occasionally have more than
50% small).

Merriman also does a 50/50 large/small breakdown (though I think he
extracts REITs first) within domestic. He also, pretty aggressively,
splits about 50/50 domestic/foreign, IIRC.

>
> Investments...nothing specific, just generally...using that many ETFs
> could lead to a lot of trading costs if you plan to round-trip buy/sell
> to rebalance the mix, or invest or withdraw smaller amounts
> periodically. Owning traditional no-load mutual funds through the fund
> company can reduce that cost.

Sometimes -- it works if you're buying all within one family and from
that family directly (ie. an all-vanguard portfolio, implemented at
Vanguard themselves).

If you're buying various no-load funds via fund supermarkets (ie. buy
Vanguard funds within a Fidelity brokerage account), you'll either get
clobbered with transaction fees, or you'll pay much higher expense
ratios (the latter for funds in the "no-transaction-fee" lists).

> Not just commissions, but also trading spreads and discount/premium to
> NAV. Also, when looking at funds vs. ETFs I have concerns about how
> ETFs will perform in less-liquid asset classes, especially when things
> get interesting.

Rick Ferri write an article about this issue particularly with respect
to corporate bond ETFs (the effect was vastly worse for junk bond funds
than for investment grade ones).

I think it's much less of an issue for most of those equity funds our
OP was talking about, but it's certainly something worth being careful
with.

I'll dig up a link if I can find it.

> I don't know how the whole ETF creation/redemption process could hold
> up in a period of high volatility or inflows/outflows, if the ETF holds
> stuff that is hard to trade (and by extension hard to value every
> millisecond).

It may do well - if lots of folks want to sell, the manager can ride it
out by turning out creation unit bundles to the market makers, and
folks who hold won't get screwed unless they, too, want to liquidate in
a panic.

Meanwhile, if folks sell open-ended funds in a panic, the whole pricing
mechanism may blow up on them because they have to redeem in cash based
on prices which may have no basis in reality, potentially screwing up
either the fund company or folks who continue to hold shares.

Vanguard may have solved the whole problem by making their ETFs a share
class of their open-ended funds.

And for highly illiquid market areas, CEFs may still make sense, at
least if they aren't overpriced (ie. high expense ratios), trading at a
premium (or even a smaller discount than "normal") and if not too
leveraged (lots of CEFs are highly leveraged, no thanks).

Anyway, the OP has, generally, selected very widely held highly liquid
ETF shares.

He may be missing out on some asset classes worth holding such as
REITs, commodities (though those can be a minefield), and even some
hedge-fund/arbitrage style assets, but by and large, it's not a
particularly wild portfolio and looks quite comparable to Merriman's
recommendations.

<http://paulmerriman.com/pauls-mutual-fund-and-etf-recommendations/>

(Note that I am *not* recommending either the OP's specific list of
funds, or Merriman's, or TILT or VTI. Just putting them out there as
points for comparison and discussion)

--
David S. Meyers, CFP�
http://www.MeyersMoney.com
disclaimer: discussions in misc.invest.financial-plan are for
educational purposes only and should not be construed as financial
advice. For personal financial advice, please consult directly with a
professional.

[Tad Borek]

On 5/1/2013 5:39 PM, Tad Borek wrote:
> Also, when looking at funds vs.
> ETFs I have concerns about how ETFs will perform in less-liquid asset
> classes, especially when things get interesting. I don't know how the
> whole ETF creation/redemption process could hold up in a period of high
> volatility or inflows/outflows, if the ETF holds stuff that is hard to
> trade (and by extension hard to value every millisecond).

As a follow-up to this thread last month...today's FT has a short piece
on how some of the market-making and creation-redemption broke down for
several ETFs yesterday. It mentions bond and emerging-market ETFs, which
I consider less-liquid asset classes. And yesterday was not exactly a
barn-burner in terms of volatility and price drops.

Bond market sell-off causes stress in $2tn ETF industry
June 21, 2013 FT.com
http://www.ft.com/cms/s/0/82d66636-d9ec-11e2-98fa-00144feab7de.html#ixzz2WrvZBR4U

Snippet:
"The selling also caused disruptions in the plumbing behind several
ETFs. Citigroup stopped accepting orders to redeem underlying assets
from ETF issuers, after one trading desk reached its allocated risk
limits. One Citi trader emailed other market participants to say: 'We
are unable to take any more redemptions today . . . a very rare
occurrence due to capital requirements we are maxed out on the amount of
collateral we have out.'"

-Tad