The Bond Dream Room

[jafs]

You could also just set them to re-balance every year.

In some ways that's the most "agnostic" approach - figuring we have no idea how things will do next year, regardless of how they've gone this year.

[dualstow]

I like both rebalancing bands and annual rebalancing, and I don't think of either as timing.
They might technically be timing, but I think of timing as, "It feels like the market is high and ready for a drop. Maybe I should sell."

[Kbg]

The literature is pretty clear...bands are better than annual. Really just think of it from a common sense viewpoint. In a flat year annual rebalancing does nothing useful and just adds cost. But if you are going to do annual, then pick May or October 1 for seasonal tendency reasons. Another potential rebalance is a clear and distinct abnormal spike...this assumes you've done your statistical homework to know what one is quantitatively. However, these are far less reliable particularly for gold.

All said and done, the best thing for "you" is to rebalance in a way that "you" can do with discipline and consistency via a set of pre-established rules. Seat of the pants rebalancing is highly likely to be both total return damaging and playing to a psychological need at exactly the wrong time for your account.

[MachineGhost]

The literature is pretty clear...bands are better than annual. Really just think of it from a common sense viewpoint. In a flat year annual rebalancing does nothing useful and just adds cost. But if you are going to do annual, then pick May or October 1 for seasonal tendency reasons. Another potential rebalance is a clear and distinct abnormal spike...this assumes you've done your statistical homework to know what one is quantitatively. However, these are far less reliable particularly for gold.

I've never seen any literature on bands that I can recall, so what was determined to be optimal? I haven't found much difference at P2T but clearly you can eek out a few bits more gain and reduce risk if you use 40/10 bands than annual rebalancing or 35/15. To me that is just exploiting the momentum effect. I wish I could backtest bands with a high level of confidence using risk parity, but I do not trust those homemade backtesters in R.

Bands really seem to be an area ripe for further exploration and experimentation, but no one seems interested because who does the PP that is a quant oriented? Almost no one. We're the Rodney Dangerfield of finance.

[Kbg]

MG,

There's definitely some out there. I've read it. I'll see if I can find a link or two. I would be surprised if band balancing wasn't better than annual with the PP. Regarding tweaking PP, I assume you've read the GestaultU stuff from a few years back. Verifying their work should be doable in AB now that it is capable of matrix operations to do the MVO analysis..my problem is A) I've long forgotten matrix Algebra and my AB custom backtest skills are about zero.

[MachineGhost]

MG,

There's definitely some out there. I've read it. I'll see if I can find a link or two. I would be surprised if band balancing wasn't better than annual with the PP. Regarding tweaking PP, I assume you've read the GestaultU stuff from a few years back. Verifying their work should be doable in AB now that it is capable of matrix operations to do the MVO analysis..my problem is A) I've long forgotten matrix Algebra and my AB custom backtest skills are about zero.

I haven't mentioned this in the Resort, but I did backtest dynamic naive risk parity in AB using a 12-month lookback period and annual rebalancing. It does outperform the regular PP a bit but is it worth all the complications over using 40/10 bands? I think programming it up to be dynamic correlated is only going to be a marginal improvement, if at all. As InsuranceGuy's research has shown, the momentum effect seems to overwhelm everything else, so I suspect that is what drives it rather than any risk parity.

We could also use some research in figuring out the ideal amount of bonds to always hedge the equity. Neither weight nor volatility seem to be a perfect match. My ideal is to completely neuter any losses. I have a hypothesis that it could be duration related.

[Kbg]

For my personal situation I am still in the asset growth phase of life so I'm more focused on growth and would (do) lever up. Today I've been messing around with portfolio analyzer using leveraged ETF risk parity. I'm going to do it more thoroughly tomorrow...it has been pretty interesting.

I still shake my head in awe at how well this particular simple mix of assets does in terms of risk/reward properties. I don't think it is particularly good for wealth building as a 1x, but that isn't a big issue to overcome.

[Kbg]

Google "band vs. annual rebalancing". Several items come up.

[curlew]

Google "band vs. annual rebalancing". Several items come up.

I generally do a combination of both - check annually and re-balance only if you've exceeded the bands, otherwise let it ride for another year unless something extraordinary is happening in the markets. I actually thought that was what HB was suggesting but I may have misunderstood. I was on a long flight back from Hong Kong when I read the book.

It's interesting that you think May or October is the best month. What is the reason for that? I generally do it late in January but it has occurred to me that if everybody else does it in January then that might be skewing things while you are in the process of rebalancing.

[Kbg]

The reason for the suggested dates is that there actually ARE seasonal tendencies in the US stock market...not always and not strong but statistically significant. Now whether they stick going forward, who knows?

However, running a quick backtest on PP using ETFs since 2005 it turns out March is the best followed by December then Jan (trade on first trading day of month)

So ignore...

[MachineGhost]

The reason for the suggested dates is that there actually ARE seasonal tendencies in the US stock market...not always and not strong but statistically significant. Now whether they stick going forward, who knows?

However, running a quick backtest on PP using ETFs since 2005 it turns out March is the best followed by December then Jan (trade on first trading day of month)

So ignore...

Yeah, it doesn't hold up on the PP probably because its so "well balanced". I use April because it is the worst month. So it's likely to be the most realistic for covering most all sitautions.

[dualstow]

TLT nearing its 52-week high today.

[dualstow]

Boy do I feel like a moron for selling my breakeven bonds that were constantly dipping into the red.
But, I'm glad I held fast to the rest. Average yield: 3.74%
Up about 31%

[MachineGhost]

Boy do I feel like a moron for selling my breakeven bonds that were constantly dipping into the red.
But, I'm glad I held fast to the rest. Average yield: 3.74%
Up about 31%

Sold my low-duration bonds and gold stocks today. Glad to get rid of that albatross. Now I can focus on pure Treasuries.

[dualstow]

Nice pop in bonds today, even though it's still early.

[barrett]

Nice pop in bonds today, even though it's still early.

Yeah, it was a good day to own LTTs. Does anyone want to take a crack at explaining to me why long bonds and gold have been so closely corellated in recent months? I haven't looked too closely at the little sqigglies, but it seems to me that those two assets have moved largely in unison the past six to eight months.

[ochotona]

I have heard it described as a rotation into risk-off assets.

[dualstow]

New Fidelity Treasury offerings announcement:

UST Maturing 07/13/2017 Auction Close Date: 04/10/2017

UST Maturing 10/12/2017 Auction Close Date: 04/10/2017

UST Maturing 04/15/2020 Auction Close Date: 04/10/2017

UST Maturing 02/15/2027 Auction Close Date: 04/11/2017

UST Maturing 02/15/2047 Auction Close Date: 04/12/2017

[dualstow]

Good post at b'heads by Robert T: The Diversifying Power of Longer Term US Treasuries

https://www.bogleheads.org/forum/viewto ... =10&t=2409

[dualstow]

I see 30-year bonds for auction via Vanguard today. Get'em while they're hot.

[Jack Jones]

I appreciate these bond auction reminders. I've missed a few auctions in the past.

[dualstow]

I appreciate these bond auction reminders. I've missed a few auctions in the past.

I meant to post it this morning -- sorry about that. Too many contractors calling, and calling about nothing, really.

Still have time, though.

[barrett]

I see 30-year bonds for auction via Vanguard today. Get'em while they're hot.

Can you explain why it's advantageous to buy bonds at auction as opposed to on the secondary market? The auctions are just something I never pay attention to. Thanks for your help.

[dualstow]

I can't. I think the last time this conversation came up, it ended with someone saying, "Some people just like that new bond smell."


EDIT: https://gyroscopicinvesting.com/forum/v ... bond+smell

[dualstow]

Slightly pro-auction comments in this TIPS thread at b'heads: https://www.bogleheads.org/forum/viewtopic.php?t=165995