IRA "Scheme"....would this be legal?

[Leon1976]

Wondering if this would be legal: ??

Take an IRA that you want to "Rothify" and divide it into two separate IRA accounts. Buy something VERY volatile...maybe like and OIL ETF with one account. Then buy the "inverse OIL" ETF in the other account (since shorting can't be done in an IRA). Then after a big move recharacterize the loser portion and re-Rothify at the lower value.

Just thinkin.....

[ddb]

Yes it's legal, but there's no perfect hedge out there (can't short, as you mention) which means that you can't be guaranteed that one will perfectly zig when the other zags. Plus, while you're waiting for the big swing you're getting a negative expected return due to the costs of the strategy.

Also, I'm not sure I see how this strategy is helpful. Let's say you buy two equal positions of $1 each, A and A'. A goes up 90%, A' goes down 90%, so now you have 1.9 in A and 0.1 in A'. Recharacterize the 0.1 in A' and now you have a Roth with 0.075 (assuming 25% conversion tax) and a Traditional worth 1.9. Total value of 1.975.

You could have just taken the original $2 and converted $0.1, giving you $1.9 in the Traditional and $0.75 in the Roth.

Am I missing something?

- DDB

[jim data]

I fail to see any benefit to convert ira's to roths. Unless you plan to withdraw tons of money and pay big taxes in retirement , i think it's better to defer the payment of tax

[smallcapvalue]

[quote="jim data"]I fail to see any benefit to convert ira's to roths. Unless you plan to withdraw tons of money and pay big taxes in retirement , i think it's better to defer the payment of tax[/quote]

Some people don't have to pay any tax now when they convert, so for them it makes sense.

[Wagnerjb]

Wondering if this would be legal: ??

Take an IRA that you want to "Rothify" and divide it into two separate IRA accounts. Buy something VERY volatile...maybe like and OIL ETF with one account. Then buy the "inverse OIL" ETF in the other account (since shorting can't be done in an IRA). Then after a big move recharacterize the loser portion and re-Rothify at the lower value.

Just thinkin.....

We discussed this exact idea at length in this post:

viewtopic.php?p=585180&highlight=pair#585180


In a nutshell, the problem with your strategy is that by investing in a "matched pair" you are guaranteeing yourself a 0% return....assuming the correlations work, as DDB so well pointed out.

Yes, you will get tax benefits, but the market doesn't have to rise by much in order for a normal equity portfolio to beat your strategy.

Best wishes.

[Tramper Al]

We discussed this exact idea at length in this post:

viewtopic.php?p=585180&highlight=pair#585180


In a nutshell, the problem with your strategy is that by investing in a "matched pair" you are guaranteeing yourself a 0% return....assuming the correlations work, as DDB so well pointed out.

Yes, you will get tax benefits, but the market doesn't have to rise by much in order for a normal equity portfolio to beat your strategy.

Or fall by much to lose either!

Supposed to be a riskless pair, though. So if you are going to stick it in your AA, as an intellectual exercise, why would you invoke an expected return on stocks? Is that a red herring or straw man? I don't really know those terms.

[EmergDoc]

I fail to see any benefit to convert ira's to roths. Unless you plan to withdraw tons of money and pay big taxes in retirement , i think it's better to defer the payment of tax

Careful not to generalize. There are lots of situations when paying the tax early is a good idea.

[Wagnerjb]

Supposed to be a riskless pair, though. So if you are going to stick it in your AA, as an intellectual exercise, why would you invoke an expected return on stocks? Is that a red herring or straw man? I don't really know those terms.

Al - I am afraid I don't understand your point. I am comparing investing $200 into the market, with investing $100 in OIL ETF and another $100 in an inverse OIL ETF. Do you feel that isn't an appropriate comparison?

Best wishes.

[xerty24]

Wondering if this would be legal: ??

Take an IRA that you want to "Rothify" and divide it into two separate IRA accounts. Buy something VERY volatile...maybe like and OIL ETF with one account. Then buy the "inverse OIL" ETF in the other account (since shorting can't be done in an IRA). Then after a big move recharacterize the loser portion and re-Rothify at the lower value.

Just thinkin.....

I did this a few years ago, using the 2x financial and -2x financial ETFs. Turns out that almost all of the tax advantage was eaten up by the net investment losses. As you may know, it's often the case that together these can lose money over a multiday period, even though they are pretty close to hedged over a single day. Saved a few bucks, but it wasn't worth the hassle and I wouldn't do that again.

[jjkthunder]

One thing that did work out for me in recharacterizing.

In 2008 I made 4 IRA to ROTH conversions. The conversions were done in April, June, Oct., and Dec.

The first 3 conversions lost a total of over $30K by the time Dec. came around.

That led me to convert the 4th conversion on Dec. 5th and on Dec 10th, I recharacterized the first 3 conversions allowing to pay tax on only the 4th conversion amount. Vanguard did all the figuring with some software formula.

I was very pleased with the recharacterization process.
.

[spam]

wagnerjb wrote:

In a nutshell, the problem with your strategy is that by investing in a "matched pair" you are guaranteeing yourself a 0% return....assuming the correlations work, as DDB so well pointed out.

To make the negative correlation work, it is easiest when both assets are in the same accont for rebalancing. If they are in seperate accounts, then it might be possible to hold cash in each account to rebalance to and from. The idea would be to mainain a constant dollar amount in cash inside each account on rebalancing day.

Rober Gibson, in his book "Asset Allocation 4th Edition" makes the best argument for the rebalancing benifit that I have seen by examining the negative correlation. All three Graphs and tables refer to the same hypothetical investment.



Graph A shows how both investments are negatively correlated and are biased for a 10% simple average return.

Table B summarizes the performance of both investments detailed in Graph A with perfectly timed annual rebalancing.

Chart C shows the outperformance of the rebalancing strategy.

The value of this analysis is to show that there is a rebalancing benifit. It would be impossible to rebalancing perfectly, but there is the opportunity for some gain. It is not simply 0%

[Wagnerjb]

Graph A shows how both investments are negatively correlated and are biased for a 10% simple average return.

Hi Spam. These "investments which are negatively correlated and are biased for a 10% simple average return" don't exist. They are like the fountain of youth in the financial world. Nice to see in a textbook, but not available in real life.

If you find one of these - that is negatively correlated to stocks, but has a nicely positive expected return - you will become fabulously wealthy. Your discovery will make bonds as obsolete as the typewriter, as people will be able to diversify AND lower volatility....without sacrificing an ounce of return.

Best wishes.

[mike_slc]

wagnerjb wrote:

In a nutshell, the problem with your strategy is that by investing in a "matched pair" you are guaranteeing yourself a 0% return....assuming the correlations work, as DDB so well pointed out.

To make the negative correlation work, it is easiest when both assets are in the same accont for rebalancing. If they are in seperate accounts, then it might be possible to hold cash in each account to rebalance to and from. The idea would be to mainain a constant dollar amount in cash inside each account on rebalancing day.

Rober Gibson, in his book "Asset Allocation 4th Edition" makes the best argument for the rebalancing benifit that I have seen by examining the negative correlation. All three Graphs and tables refer to the same hypothetical investment.

In his example, investments C and D both have 10% expected return. With an short pairing or a perfect inverse ETF, the expected return of the short or inverse ETF is the inverse of the expected return of the original investment.

Example: Investment A has expected return 10%. Short A or perfect inverse ETF of A has expected return of -10%. Thus pairing them together should dampen volatility to almost nothing and give you a return of... 0%. Less than that actually because of costs.

[spam]

wagnerjb wrote:

In a nutshell, the problem with your strategy is that by investing in a "matched pair" you are guaranteeing yourself a 0% return....assuming the correlations work, as DDB so well pointed out.

To make the negative correlation work, it is easiest when both assets are in the same accont for rebalancing. If they are in seperate accounts, then it might be possible to hold cash in each account to rebalance to and from. The idea would be to mainain a constant dollar amount in cash inside each account on rebalancing day.

Rober Gibson, in his book "Asset Allocation 4th Edition" makes the best argument for the rebalancing benifit that I have seen by examining the negative correlation. All three Graphs and tables refer to the same hypothetical investment.

In his example, investments C and D both have 10% expected return. With an short pairing or a perfect inverse ETF, the expected return of the short or inverse ETF is the inverse of the expected return of the original investment.

Example: Investment A has expected return 10%. Short A or perfect inverse ETF of A has expected return of -10%. Thus pairing them together should dampen volatility to almost nothing and give you a return of... 0%. Less than that actually because of costs.

Hi Mike,

Look at the rest of the page. The compound annual return for each investment by itself is 5.83% where rebalancing both yields a compound annual return of 10%.

[spam]

Graph A shows how both investments are negatively correlated and are biased for a 10% simple average return.

Hi Spam. These "investments which are negatively correlated and are biased for a 10% simple average return" don't exist. They are like the fountain of youth in the financial world. Nice to see in a textbook, but not available in real life.

If you find one of these - that is negatively correlated to stocks, but has a nicely positive expected return - you will become fabulously wealthy. Your discovery will make bonds as obsolete as the typewriter, as people will be able to diversify AND lower volatility....without sacrificing an ounce of return.

Best wishes.

Hi wagnerjb,

The O.P. suggested choosing something volitile and then shorting it. While finding something in the real world to exactly match the hypothetical would be difficult to impossible, it does show what the best scenario would be and provides a nice intuitive visual idea.

Here is a real life example of a negative correlation using mutual funds.



With the S&P 500 returning less than zero over the past decade, the strategy of shorting combined with rebalancing would certainly have improved it.

[Wagnerjb]

The O.P. suggested choosing something volitile and then shorting it. While finding something in the real world to exactly match the hypothetical would be difficult to impossible, it does show what the best scenario would be and provides a nice intuitive visual idea.

Hi Spam: I understand shorting, and I agree that shorting will give you the negative correlation you want. But shorting (or buying an inverse fund) will yield a total return of 0% for the two investments combined. The notion that you can come up with two investments with negative correlation AND a combined positive return is what I wanted to disagree with.

Once you accept that the two investments combined will yield 0%, the tax benefits suddenly look less attractive.

Best wishes.

[spam]

The O.P. suggested choosing something volitile and then shorting it. While finding something in the real world to exactly match the hypothetical would be difficult to impossible, it does show what the best scenario would be and provides a nice intuitive visual idea.

Hi Spam: I understand shorting, and I agree that shorting will give you the negative correlation you want. But shorting (or buying an inverse fund) will yield a total return of 0% for the two investments combined. The notion that you can come up with two investments with negative correlation AND a combined positive return is what I wanted to disagree with.

Once you accept that the two investments combined will yield 0%, the tax benefits suddenly look less attractive.

Best wishes.

Hi Andy,

Yes, we are in disagreement on the same point. I am suggesting that Gibson does a fine job of demonstrating the rebalancing benifit using negatively correlated assets.

If you simply buy and hold without rebalancing, then yes the simple average gain would be 0%.

However, if you are rebalancing, then you are selling shares of asset "A" when they are high priced, and buying shares of "B" when they are low priced. If both A and B changed in price between $5 and $10, then you would be selling 1 share of "A" at $10 and buying 2 shares of "B" at $5. Next year, you would be selling 2 shares of "B" at $10 (for example) and buying 4 shares of "A"at $5.

This is the rebalancing benifit that Gibson is illustrating in his perfect world example.

If the picture I posted does not display, try pressing F5 (for windows) to refresh the page.

[Wagnerjb]

If you simply buy and hold without rebalancing, then yes the simple average gain would be 0%.

However, if you are rebalancing, then you are selling shares of asset "A" when they are high priced, and buying shares of "B" when they are low priced. If both A and B changed in price between $5 and $10, then you would be selling 1 share of "A" at $10 and buying 2 shares of "B" at $5. Next year, you would be selling 2 shares of "B" at $10 (for example) and buying 4 shares of "A"at $5.

This is the rebalancing benifit that Gibson is illustrating in his perfect world example.

OK, I see the point...but maybe not how it applies to this situation. Yes, we all get a rebalancing benefit to the extent our asset classes are not highly correlated, and perfectly negative correlation would provide great rebalancing benefits.

But, don't you agree that the Roth conversion example doesn't lend itself to this example (if we assume the tooth fairy existed and brought us negative correlation as a gift)? You need time for the rebalancing to bear fruit, and you don't have that in this case.

In the short space you have, the inverse fund (or shorting) would provide the negative correlation. But you would have zero return....unless as you say you could rebalance multiple times.

Best wishes.

[spam]

Warnerjb wrote:

OK, I see the point...but maybe not how it applies to this situation. Yes, we all get a rebalancing benefit to the extent our asset classes are not highly correlated, and perfectly negative correlation would provide great rebalancing benefits.

But, don't you agree that the Roth conversion example doesn't lend itself to this example (if we assume the tooth fairy existed and brought us negative correlation as a gift)? You need time for the rebalancing to bear fruit, and you don't have that in this case.

In the short space you have, the inverse fund (or shorting) would provide the negative correlation. But you would have zero return....unless as you say you could rebalance multiple times.

Best wishes.

Hi Andy,

Yes, I do agree. The best (or easiest) way to benifit from a negative correlation would be to have both assets in the same account. What the O.P. proposes does not lend itself well to rebalancing therefore the 0% gain is a real possibility.

The only way I can think to get around this would be to hold cash in both accounts. Shares of the outperforming fund would be sold to cash, and shares of the underperforming fund (in a different account) would be bought from cash at the same time.

This would be less than ideal unless the cash position was considered to be part of the fixed income allocation.

[Kingsroad]

Wondering if this would be legal: ??

Take an IRA that you want to "Rothify" and divide it into two separate IRA accounts. Buy something VERY volatile...maybe like and OIL ETF with one account. Then buy the "inverse OIL" ETF in the other account (since shorting can't be done in an IRA). Then after a big move recharacterize the loser portion and re-Rothify at the lower value.

Just thinkin.....

In the most simple example, you'd put half of your IRA into a Roth "bet" on a coin toss covering heads, while your other Roth is used to cover tails on the same coin toss. A few seconds after the coin leaves your hand, and gravity takes over, one of your Roth's will have lost 100% while the other doubled. You could send one back and keep the winner, reducing your tax cost by 50%.

One problem: Economic Benefit Doctrine would likely allow the service to disallow the recharaterization since there is no profit motive in the transaction, only tax savings.

However, properly structured option contracts can imitate the coin toss result (not perfectly because of the transaction costs and time value decay), with the benefit that if the market has sufficient movement, one of the "bets" could pay more than even money, making the overall transaction profitable without the tax savings.

If you'd like to discuss particulars, check your message box.

[Wagnerjb]

However, properly structured option contracts can imitate the coin toss result (not perfectly because of the transaction costs and time value decay), with the benefit that if the market has sufficient movement, one of the "bets" could pay more than even money, making the overall transaction profitable without the tax savings.

Do you care to tell the rest of us how to get this free lunch? So, using options you can essentially duplicate the coin toss (one total loss, one doubles in value).....but with a total pretax return that is greater than zero? Forget the Roth conversion idea, let's use your strategy in our regular accounts. Let's see....zero investment risk, positive return. What's not to like?

Please explain more. (I am assuming you have a legitimate strategy in mind and that you are not fishing for customers at this website.)

Best wishes.

[Kingsroad]

However, properly structured option contracts can imitate the coin toss result (not perfectly because of the transaction costs and time value decay), with the benefit that if the market has sufficient movement, one of the "bets" could pay more than even money, making the overall transaction profitable without the tax savings.

Do you care to tell the rest of us how to get this free lunch? So, using options you can essentially duplicate the coin toss (one total loss, one doubles in value).....but with a total pretax return that is greater than zero? Forget the Roth conversion idea, let's use your strategy in our regular accounts. Let's see....zero investment risk, positive return. What's not to like?

Please explain more. (I am assuming you have a legitimate strategy in mind and that you are not fishing for customers at this website.)

Best wishes.

My post indicates that there are transaction costs and time decay that would not be risk free.