What's Next For Equity Markets? And What To Do?

[Moemg]

Audrey , I am always impressed with your knowledge of managing your portfolio especially the tax consequences . Were you a CPA before you retired ? If not how did you get accumulate your knowledge ?

 

[audreyh1]

Audrey , I am always impressed with your knowledge of managing your portfolio especially the tax consequences . Were you a CPA before you retired ? If not how did you get accumulate your knowledge ?

Nope! Engineer (computer hardware and software). My husband became somewhat of a tax expert when he started his own consulting business (also an engineer). He was willing to crack the books and figure out Turbotax in detail. He still does the annual taxes, but since I manage all our investments I took over the estimated taxes when I retired. That required me to learn what I needed to know. The IRS publications pretty much spell things out so it's more a matter of reading them than anything. Reading and doing!

In terms of tax consequences of the portfolio - that's simple. 95% of our investments are in taxable accounts. Every year since retirement I have tracked the % of the portfolio that was paid in taxes that year since it comes out of our withdrawal. This has allowed me to observe and learn from the various trends. Morningstar also rates the tax efficiency for mutual funds and that can provide some insight. That helps somewhat.

Unfortunately it's the riskier (most volatile) investments (equities with no dividends - growth stocks and small caps) that are the most tax efficient.

Audrey

 

[freebird5825]

...I have small positions (both formerly at least 2X) in REIT (VGSIX) and Comm/Energy (VGENX) in tax sheltered accounts that I wonder about in terms of relative upside right about now. I believe VGENX does have a lot of upside, but it's hard to imagine any REIT coming back strong any time soon.

I also have small stakes in both, tucked away in a Roth IRA.
I knew they were going to be "bucking broncos", but never anticipated this bad a result. Live and learn.
I can't contribute any more to the Roth until a) I have earned income...no way , or b) get married...not until 2013.
So I am just letting them simmer.
I do like the dividends VGSIX is pumping out , only as a small consolation to 2008's damage. Ouch!

 

[Moemg]

Nope! Engineer (computer hardware and software). My husband became somewhat of a tax expert when he started his own consulting business (also an engineer). He was willing to crack the books and figure out Turbotax in detail. He still does the annual taxes, but since I manage all our investments I took over the estimated taxes when I retired. That required me to learn what I needed to know. The IRS publications pretty much spell things out so it's more a matter of reading them than anything. Reading and doing!

In terms of tax consequences of the portfolio - that's simple. 95% of our investments are in taxable accounts. Every year since retirement I have tracked the % of the portfolio that was paid in taxes that year since it comes out of our withdrawal. This has allowed me to observe and learn from the various trends. Morningstar also rates the tax efficiency for mutual funds and that can provide some insight. That helps somewhat.

Unfortunately it's the riskier (most volatile) investments (equities with no dividends - growth stocks and small caps) that are the most tax efficient.

Audrey

Thanks Audrey ! This is one subject I really need to learn more about so I will hit the books .

 

[tulak]

Thanks Haha. This is a wonderful thread, one of the best I've read here in awhile (IMO of course). I'm primarily a lurker, but I'm delurking in this case, since this touches on many points I've been thinking about. And it looks like I'm not the only one...

Market timing doesn't work for those who do not have inside information. This is not new information on this board, or I would provide links.

Decide on your asset allocation, diversify, and hang on for the ride.

(If you have money to invest, dollar cost average, dollar value average, or don't. I like to DCA because it is easy and because that way I don't kick myself quite so much.)

I wonder about this. What if you chose to change your allocation based on the valuation of an individual fund, using PE? For example, let's say you want to have 25% Total Market at PE 15. For each change in PE, change by a specified %, let's say 1%. So at PE 10, you'd be at 30% and at PE 20 you'd be at 20%.

The question I haven't answered: is this better or worse than just riding it out with a fixed allocation? But even if it was a wash in dollars, I think you'd almost certainly have less downside risk since you're decreasing your total exposure.

Again, this is a just a thought at this point. I haven't had the time to fully investigate this idea. But maybe somebody on the board is already aware of the outcome.

Plenty has been written about buying, and relatively little about strategies for selling (perhaps this is the selfish bias of Wall Street in action). One thought is to follow Audrey's example and value average down by reducing my exposure by say 2% per year (larger than her 1% since I'm starting from a higher equity allocation).

When the downturn occurred, I started putting more thought in how to achieve a more conservative allocation. I always planned on a 60/40 allocation at retirement, but never had a concrete plan on how to get there.

I've been holding at 90/10 throughout the downturn, even though the feeling of buying funds that are going down is difficult. On another thread, I read it makes you feel downright dumb. I believe that's an accurate description.

What I came up with is to decrease my equity holdings every year by 2%. This results in having a 60/40 allocation at 50 and 40/60 at 60. A bit more conservative than I originally planned for, but at that point, I really don't need the extra risk.

Over the last week, I've dollar value averaged to 88/12.

I'm still intrigued by the idea of changing allocation based on valuation, but until I come up with a concrete plan on how it would work, I'm sticking with my original plan...

 

[audreyh1]

Thanks Audrey ! This is one subject I really need to learn more about so I will hit the books .

Pretty much Instructions for Form 1040ES and Publication 505 covers what you need to know about estimated taxes. All available as downloads from irs.gov

Audrey

 

[ziggy29]

Nope! Engineer (computer hardware and software). My husband became somewhat of a tax expert when he started his own consulting business (also an engineer).

Nice to know at least some software engineers got an early-out. I thought mine was coming until last year....

 

[audreyh1]

Well a few of us engineer types got real lucky in the late 1990s even though we had some severe market events in 1997 and 1998!

Audrey

 

[Bikerdude]

I wonder about this. What if you chose to change your allocation based on the valuation of an individual fund, using PE? For example, let's say you want to have 25% Total Market at PE 15. For each change in PE, change by a specified %, let's say 1%. So at PE 10, you'd be at 30% and at PE 20 you'd be at 20%.

I would think this is what you accomplish when you rebalance.

 

[tulak]

I would think this is what you accomplish when you rebalance.

It's actually a little bit more aggressive than a standard rebalance.

For example, let's say you 50% in TSM and 50% in FI. TSM has PE15. TSM/FI goes to 55/45 and the TSM PE increases to 20. With standard rebalancing, you'd go to 50/50, but now your TSM has PE20.

With what I'm thinking, you'd take one more step. If you decided on a +/- 1% allocation change per equivalent change in PE, you would have changed your allocation from 50/50 to 45/55, taking into account a +5 change in PE.

As I said, this is an idea that I've recently been thinking about. I have no idea if it's practical, makes no sense, etc. But I know I'd appreciate some method I can use to take valuations into account.

 

[freebird5825]

...As I said, this is an idea that I've recently been thinking about. I have no idea if it's practical, makes no sense, etc. But I know I'd appreciate some method I can use to take valuations into account.

If I understand correctly what it is you are trying to do...are you trying to figure out how to rebalance your portfolio to a certain AA based on its current price per share or NAV?
If I am misunderstanding, then disregard what follows.

Warning...techno-babble alert!

I essentially use a simple raw numerical reduction method by varying one or more concurrent variables (investments) to reach a desired end state (target AA). The deltas are the number of shares and/or addition (buy) or reduction/removal (sell) of a given investment.
My algorithm is exercised using an online portfolio modeler tool. I use M* to cover stocks, bonds, cash, and mutual funds.
I enter all ticker symbols & number of shares actually owned, run the XRay tool analyzer on it to get the current AA as well as other measures of interest to me.
Next step is to make a copy of that portfolio (store the baseline) and play around with the share numbers and holdings in the copy.
I use a share number = 1 if I am modelling a complete sell off but don't want to keep typing in the ticker symbol over and over...
Rerunning the AA analyzer on the copy tells me which holdings to buy or sell to achieve the new and improved AA as well as a measure of the diversification across sectors. The M* Stock Intersection tool is invaluable here.
Having the original stored gives me a look-back for comparisons or return point for individual holdings.

I do this at Morningstar all the time to model planned changes. Then I track it for a while before I actually make a move to exchange/sell/buy.

It is not an exact science, since the datum (price per share or NAV) are constantly moving with daily market fluctuations, but it is my best approximation.

 

[CCdaCE]

It's actually a little bit more aggressive than a standard rebalance.

For example, let's say you 50% in TSM and 50% in FI. TSM has PE15. TSM/FI goes to 55/45 and the TSM PE increases to 20. With standard rebalancing, you'd go to 50/50, but now your TSM has PE20.

With what I'm thinking, you'd take one more step. If you decided on a +/- 1% allocation change per equivalent change in PE, you would have changed your allocation from 50/50 to 45/55, taking into account a +5 change in PE.

As I said, this is an idea that I've recently been thinking about. I have no idea if it's practical, makes no sense, etc. But I know I'd appreciate some method I can use to take valuations into account.

What makes you think PE has any predictive power and/or should be used as an asset allocation tool? Why not dividend yield equivalent or Tobin's q?

How do you define PE? Using past earnings, future earnings, etc.?

-CC

 

[tulak]

What makes you think PE has any predictive power and/or should be used as an asset allocation tool? Why not dividend yield equivalent or Tobin's q?

How do you define PE? Using past earnings, future earnings, etc.?

-CC

I don't know if it has predictive power and I don't know if it should be used as a tool to determine AA. But I am thinking about how I might be able to use this in my AA.

Now that I had a little bit more time, I searched on dividend yield equivalent, Tobin's q and read a little about Shiller's PE10.

Not surprisingly, others have also thought about this idea.

I'd say what I'm proposing probably aligns closest to Shiller's PE10 in terms of determing valuations.

In searching, I've found articles such as this:

Is P/E ratio a useful stock valuation measure? « The Investment Scientist

To answer this question, I examined the whole stock market data for the past 50 years from 1958 to 2007. For each year, I separated stocks into three portfolios: the top 30% P/E portfolio, the middle 40% P/E portfolio and the bottom 30% P/E portfolio. (Stocks with negative earnings are all in the top 30% P/E portfolio.)
If I had invested $1 in each of the three portfolios at the beginning of 1958, by the end of 2007, the top 30% P/E portfolio would have grown to $91; the middle 40% P/E portfolio would have grown to $322 and the bottom 30% P/E portfolio would have grown to $1698!

That leads me to believe there are merits to integrating some type of valuation within an AA.

Keep in mind, although the quoted text above shows a $ gain investing in lower PE assets, that's not my goal. My goal is reduce risk (volatility) by investing in assets with lower valuations and selling assets with high valuations.

At some point (hopefully soon) I'll model this method and see if it makes any sense within my AA...

 

[Grep]

If the S&P makes it back to 1000 anytime soon, this should be enough to trigger my out-of-balance even with my new wider "tolerance band".

Audrey,

I'm curious about your strategy for rebalancing in taxable at that point. I'd assume that you would have some shares underwater, some about even, and some up, plus potentially some substantial paper losses due to tax-loss-harvesting. So what would you sell, in what order, and with what rationale?

(Others should feel free to chip in too.)

Thank you.

 

[DblDoc]

I can't speak for Audrey but in my own case I have TSM in both a taxable account and in tax deferred so I am able to rebalance without creating a taxable event.

DD

 

[Grep]

Yes, in tax-deferred, there's no point to a basis (from a tax standpoint) and it don't matter what you do. The question is about taxable accounts.

Thanks, though.

 

[audreyh1]

Audrey,

I'm curious about your strategy for rebalancing in taxable at that point. I'd assume that you would have some shares underwater, some about even, and some up, plus potentially some substantial paper losses due to tax-loss-harvesting. So what would you sell, in what order, and with what rationale?

(Others should feel free to chip in too.)

Thank you.

Rebalancing (in a non tax-deferred account) is a taxable event. Can't really do anything about it. If I'm trimming winning equity funds, I'm realizing gains so there is a tax cost - but at least it's at capital gains tax rates. If equities are down and I'm selling bond funds to buy equities that is often a tax loss event because I each time I pay taxes on bond fund distributions the fund basis is raised.

I don't choose anything according to its basis - each fund is trimmed or added to by how much it is above or below its target allocation.

Audrey

 

[audreyh1]

Oh, wait - there"s more......!

An important little refinement that attempts to minimize the taxable events due to rebalancing:

One trick I employ is to not reinvest distributions in mutual funds that are have moved above of their target allocation during the year since I have to pay taxes on these distributions anyway. This takes care of some of the "trimming".

Audrey

 

[Grep]

If equities are down and I'm selling bond funds to buy equities that is often a tax loss event because I each time I pay taxes on bond fund distributions the fund basis is raised.

Thanks Audrey,

Please pardon what may be elementary questions.

First, can you explain the above point further? How do bond fund distributions (by this you mean dividends or sales of shares?) increase the basis?

Then, suppose that through market turmoil, some (stock) shares are up, and some down, yet you are intent on selling some shares. Which shares go, and in what order, assuming you use the specific shares method of reporting?

Then, suppose you also have a realized loss through tax loss harvesting, well in excess of what you expect to use via a yearly 3k deduction. How might this affect your decision?

Thank you again.

 

[audreyh1]

Distributions (mutual fund distributions include dividends or any short and long-term capital gains realized ) paid from a mutual fund increase the basis because you pay tax on the distributions. That's how taxes on mutual funds work. Then when you sell mutual fund shares later, any distributions paid out already increase the basis because you already paid tax, you only pay tax on any additional gains.

On picking which shares of a stock to sell during rebalancing — well, I only own mutual funds and I use the average cost basis, so I don't get to pick a specific lot to sell. Fidelity tracks this average basis for me (and the calculation takes into account the raised basis due to all distributions paid). Of course if you DO get to select a specific lot then choosing the lot with highest basis first held for at least 12 months would be prudent. Easy enough to do with stocks, but IMO too complicated for me with mutual funds.

If you have realized a loss via tax loss harvesting, this can be applied against any (taxable) capital gains distributions paid out by mutual funds that same year. Only once you've exceeded those are you then limited to $3K against ordinary income.

Audrey

 

[Grep]

Thanks Audrey,

Ah, yes, now I dimly remember that basis issue (hasn't come up for me yet).

I've been using the specific shares method, with manual book-keeping. I don't automatically reinvest dividends, etc., to keep that more manageable. Instead, I've used that money to make less frequent purchases for rebalancing, etc.

The general advice is to sell the highest-basis shares first, as that normally results in the least tax bite. However, if you have a mix of shares, some that are up and some down, selling the highest basis shares results in being taxed, which is avoidable.

Alternatives include (a) selling what shares you have whose basis is (approximately) fairly valued, for no substantial net gain or loss, or (b) selling a mix of the highest basis and lowest basis stock such that there is no net gain or loss. I think these two are completely equivalent, but I should do some math to verify this.

Another option is (c) to take the gain (sell lowest basis) and run, to the degree that old excess tax-harvested losses are available. However, this is effectively paying back the "tax-free loan" aspect of tax-loss harvesting, which one would normally like to stretch out.

Finally, one could sell the highest basis shares for a loss. However, this wouldn't seem to make much sense except for further tax-loss harvesting.

So, I think (a) and (b) are equivalent and probably is the best strategy in the absence of some reason why taking gains against old losses (c) is preferable. Perhaps changes in tax rates and rules might trigger the latter.

(Edited) Now I think that perhaps a more optimum strategy is to sell the fewest number of shares possible, however they are allocated, consistent with any other goals such as avoiding paying capital gains taxes. This maximizes future dividend payments. However, this may also be equivalent with (a) and (b).

Any comments from anyone, or links to discussions of this sort of optimization?

Thanks.

 

[NW-Bound]

The following is excerpt from IRS Publication 564, regarding various methods to determine the cost basis. As it is usually of a tax advantage for the investor to sell the most recently acquired shares first, in contrast to the early shares which tend to have a lower basis, the IRS has provided more hurdle to the share seller. He would have to designate the shares identified by date of purchase to his broker, who then has to provide confirmation in writing.

I know of no online brokers who have software to let you make such designation to sell. And if the shares have been transferred from one brokerage to another, the purchase history is lost, and is only known by the investor who hopefully has kept a paper record. He still has no way of telling his new broker to sell XX shares purchased on a certain date with the previous broker.

Because of this hassle, I use the FIFO method, which is of course not tax-efficient. However, I have more stocks than mutual funds, and usually do tax loss selling on the stocks in the taxable accounts. If I need to sell mutual funds for AA rebalancing, I usually sell MFs in the tax-deferred accounts to avoid the basis computation hassle.

Cost Basis

You can figure your gain or loss using a cost basis only if you did not previously use an average basis for a sale, exchange, or redemption of other shares in the same mutual fund.

To figure cost basis, you can choose one of the following methods.

Specific share identification.

First-in first-out (FIFO).
​

Specific share identification. If you adequately identify the shares you sold, you can use the adjusted basis of those particular shares to figure your gain or loss.
You will adequately identify your mutual fund shares, even if you bought the shares in different lots at various prices and times, if you:

Specify to your broker or other agent the particular shares to be sold or transferred at the time of the sale or transfer, and
​

Receive confirmation in writing from your broker or other agent within a reasonable time of your specification of the particular shares sold or transferred.​

You continue to have the burden of proving your basis in the specified shares at the time of sale or transfer.

First-in first-out (FIFO). If your shares were acquired at different times or at different prices and you cannot identify which shares you sold, use the basis of the shares you acquired first as the basis of the shares sold. In other words, the oldest shares you own are considered sold first. You should keep a separate record of each purchase and any dispositions of the shares until all shares purchased at the same time have been disposed of completely.
Table 3 (on the next page) illustrates the use of the FIFO method to figure the cost basis of shares sold, compared with the use of the single-category method to figure average basis (discussed next).

Average Basis

You can figure your gain or loss using an average basis only if you acquired the shares at various times and prices, and you left the shares on deposit in an account handled by a custodian or agent who acquires or redeems those shares.

To figure average basis, you can use one of the following methods.

Single-category method.

Double-category method.​

Once you elect to use an average basis, you must continue to use it for all accounts in the same fund. (You must also continue to use the same method.) However, you may use the cost basis (or a different method of figuring the average basis) for shares in other funds, even those within the same family of funds.

Single-category method. Under the single-category method, you find the average basis of all shares owned at the time of each disposition, regardless of how long you owned them. Include shares acquired with reinvested dividends or capital gain distributions.
Table 3 illustrates the use of the single-category method to figure the average basis of shares sold, compared with the use of the FIFO method to figure cost basis (discussed earlier).
Even though you include all unsold shares of a fund in a single category to compute average basis, you may have both short-term and long-term gains or losses when you sell these shares. To determine your holding period, the shares disposed of are considered to be those acquired first.

Double-category method. In the double-category method, all shares in an account at the time of each disposition are divided into two categories: short term and long term. Shares held 1 year or less are short term. Shares held longer than 1 year are long term.
The basis of each share in a category is the average basis for that category. This is the total remaining basis of all shares in that category at the time of disposition divided by the total shares in the category at that time. To use this method, you specify, to the custodian or agent handling your account, from which category the shares are to be sold or transferred. The custodian or agent must confirm in writing your specification. If you do not specify or receive confirmation, you must first charge the shares sold against the long-term category and then charge any remaining shares sold against the short-term category.
​

 

[samclem]

The general advice is to sell the highest-basis shares first, as that normally results in the least tax bite. However, if you have a mix of shares, some that are up and some down, selling the highest basis shares results in being taxed, which is avoidable.

?? The shares with the "highest basis" are the shares that you paid the most for. Since you pay tax on the difference between the price you paid when you bought them and the price when you sold them, these shares would generate the lowest tax bite.

Part of any optimization strategy will involve guessing the the expected future tax rate you will have to pay. If (as is presently scheduled to happen) the Cap gains rate goes up to match the earned income rate, you might have been better off to pay the cap gains now rather than the higher rate later.

I'm no expert. I use the average cost basis and keep things simple, though I know it costs more money in taxes.

 

[clifp]

With the huge caveat if will past IRS muster or not... Here is what I do.

Years ago I sent a letter to Schwab instructing them to sell the stock with the highest cost basis for gains, and the lowest cost basis for losses.

Over the last 5 or 6 years it has actually gotten easier to implement. In the performance section of the website, there is a way of entering cost basis (generally it is filled in automatically) for shares. Then when you sell the shares Schwab ask you to identify which shares you sold so I fill that section out.

Now, this method doesn't live up to the letter of the IRS regulation, but since I do identify the specific shares I am selling shortly after the sales (and generally before the trade clears), I think I am living up to the spirit. There is also an electronic audit trail. If the IRS wants to come after me so be it, but I've been doing this for well over a decade with no problems.

 

[Grep]

NW-Bound:

Vanguard will manually send you the "adequate" documentation identifying specific shares if you call them up (or at least they will if you are a higher-tier customer). Once you take care of getting the documentation, you go ahead and sell on your own the normal way. It's a huge pain compared to merely clicking a few buttons, though.

Samclem:

Correct - I mis-wrote. I need to hurry up and retire before my last few functioning neurons go.

I remember earlier discussions (pre-crash) about the expected rise in capital-gains taxes, and people doing break-even analysis. Post-crash, it seems that many fewer people need worry about capital gains. But who knows, maybe there will be a mad dash to a full recovery just in time.

Clifp:

I'm with you on the "sprit" of the rules.

Vanguard should get up to speed and track specific shares in mutual funds. Meanwhile, the IRS is obviously still stuck in an earlier age given their expectation that a "broker" is involved.

By the way, there are those who say that one can informally switch from FIFO to Specific Shares, as they are compatible methods (the specific shares are just the earliest) provided that one never specifies FIFO (FIFO is the default, though) and finally does specify Specific Shares once that method is employed. I'm not sure I'd care to try that, though, as the IRS could conceivably claim that you have switched methods without permission.

Thanks.