Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
Probably wise. Got tax implications to consider... But it makes DSENX/DSEEX's performance all the more impressive, presuming that it would avoid tech as overvalued. Anyway to see how it's currently allocated? It's hard to figure out what all the derivatives mean.
An interesting Link with some terrific charts. But no great surprise given that the S&P 500 is a capitalization weighted index.
That weighting formula will move any correlation coefficient higher when the overall index is compared to the top 1, or 5, or 10 of its top holdings.
If equally weighted any firm would have a minor impact of that Imdexes movements. However, with the cap weighting operating, the top 5 firms currently make up over 10% of its daily movements. Here is a Link to Morningstar that documents the current weighting distribution for the top current S&P 500 index positions:
I don't know if that weighting has changed dramatically recently, but I'm sure it has indeed changed. That's what makes the marketplace so uncertain and risky.
Using Portfolio Visualizer the chart in David's article doesn't look all that much different when using the S&P 500 (I used VOO) in comparison to the FAANG stocks. Here what PV produced. I used equal weight for the FAANG stocks. (FAANG portfolio is in Blue, VOO in Red):
What I find interesting is that the FAANG stocks, as a stand alone portfolio, have only a .52 correlation to the market (VOO is almost perfectly correlated at .99). So the FAANG stocks are adding alpha to the S&P 500 index in an uncorrelated way. My question is if FAANG's out performance is uncorrelated to the S&P 500 performance will FAANG's under performance be less of a concern? Said another way, if the FAANG stock perfromance has not caused the rest of the S&P 500 to also "catch fire" why would it's under performance cause the S&P 500 to "crash and burn"?
I believe one could argue that over investment in just a few stocks may suppress investment in the others.
One quick note about DSENX that seemed curious. DSENX has exhibited consistently deeper drawdowns (one could argue only slightly deeper) than VOO (Vanguard S&P 500) shown here (DSENX is in Red):
Unfortunately, as of last month, all 5 of those stocks were in DSENX because the four sectors were technology and consumer cyclical, which cover those 5, healthcare and industrials. Doubleline hasn't provided information about this month's sectors on their website yet.
The derivatives aren't that complicated. If you consider taking a portfolio of $3BN and investing 25% in each of the 4 sectors, and I think it's based on SPDR's Select Sector etfs, then they're doing exactly the same thing. They're just doing it as a "paper bet" with Barclay's where they "settle" the bet at a defined date in the future by the loser paying the winner rather than actually buying and selling all the stocks along the way. Since they still have all the cash in their hands until they have to settle the bet, they invest all the cash in an actively managed collection of bonds. Those bonds offset the cost they pay for making the paper bet, the expense ratio and historically have provided additional income to the fund.
I'm not a tax expert but I think it's interesting that a decent portion of the fund's distributions for the last few years have been long-term gains. My general understanding is that the gains and losses from the swaps should be ordinary income so I would love to understand how they're distributing the long-term gains that they are.
You also can just track RSP since its May 03 inception and can get a sense of how it has outperformed SP500 (from 14y down to ~4y) and then underperformed SP500 (the last 3-4y).
While CAPE has trounced both over its 4.7y life.
Interesting the degree to which RSP deviates from true equal weighting:
@LLJB "Unfortunately, as of last month, all 5 of those stocks were in DSENX". That could explain bad results for the fund last days. However, DSENX is supposed to invest in value stocks, while all FAANG are considered growth ones. Can you please provide a link.
Its reasonable to think they're investing in value stocks but don't forget their definition is different than what, M* for instance, calls a value stock. CAPE looks at the price now compared to normalized earnings over 10 years.
Here's a link to a white paper about how they construct the index they're following.
I haven't read and thought in depth about every detail but one very important detail of their "relative CAPE ratio" is that it looks at a sectors CAPE ratio relative to what has been normal for that sector. If I understand correctly, then technology, which might have had a higher CAPE ratio historically because of higher growth, can still be selected when it's current CAPE ratio is higher than other sectors but not high compared to the history for that sector. The discussion starts on page 19 of the paper (recognizing that for some reason the paper starts on page 16).
EDIT: I had posted a comment about the impact of market cap that I've thought more about and decided was incorrect so I've deleted it and added a bit at the end.
If I have this right then whether you end up investing in value or growth, not just because of the difference between current CAPE and current P/E, depends on whether the stocks with the highest market cap are "relative CAPE" based cheap or expensive compared to their own sector. That makes expectations for a "value" investment more questionable and could explain some of the performance history. I don't know the history well enough to make any judgment and apparently that history isn't easy to find. I went back through Doubleline's historical SEC filings and they only started disclosing (in their SEC filings) the sectors that the fund was invested in a little more than a year ago. Technology was in the fund as of March 2016 as well. Maybe it would be possible to find historical information based on Barclay's index but I don't have time at the moment to search for it.
One data point that might not be good enough for drawing conclusions but provides a little perspective is the current investments. The traditional CAPE ratio says Real Estate is the most overvalued now, then Technology, Communication Services (which I think Barclays combines with Technology), Healthcare, Basic Materials, Consumer Cyclical, Utilities, Industrials, Consumer Defensive, Financial Services and Energy. The fund takes the 5 least expensive and picks the 4 with the best momentum, so tech, healthcare and consumer cyclical wouldn't even make it into the momentum analysis based on traditional CAPE but they all make it based on "relative CAPE".
@LLJB super thanks for these insights. I'd just glanced at M*'s page and couldn't understand what the derivatives meant for sector breakdown; I also had simply assumed that since this fund is supposed to have a value tilt, it wouldn't be in the FAANGs.
Your speculations about what it's doing make sense. I'm in it and obviously pleased so far but it's one I'd probably sell rather than add to (unlike say, TDVFX) if it starts underperforming.
@expatsp, yes, I'm in it and I think what I've been learning just makes me want to control my exposure more than I was pondering previously.
TDVFX is a fund I like a lot and I'm not in it yet but I will be eventually even if I have to open an account directly with them.
@davidrmoran, yes, you're right, but in determining how expensive each sector is they use a modified CAPE ratio that causes different results. If I understand correctly they've taken the traditional CAPE and divided it by the historical average CAPE (I believe over 20 years). That causes sectors like technology and healthcare, which both have high CAPE ratios now relative to the other sectors, to be considered "cheap" because on average their CAPE ratios have been higher than other sectors for a long time. Once they get to their version of least expensive then they indeed take the 4 with the best momentum out of the least expensive 5 as you said.
Thank you so much. And here I was thinking DSENX was somehow going to protect me in an S&P 500 carnage. I added it in a portfolio I was not supposed to sell out of, but now I know it should be a sell candidate per my ANALysis.
I really feel like an idiot. I should have been monitoring DSENX closely.
So is this the beginning of a time of testing for CAPE?
Ten days ago it diverged significantly downward from SP500 quality indexes: SPHQ, SCHDH, DVY, NOBL, OUSA, and similar. RPS is in the middle of these since midweek last week.
What does this signify --- that CAPE occasionally does get FANGed?
At least we'll see if and how it adjusts. I'd guess energy is the sector that gets eliminated because of no momentum, and I can't imagine technology is going to lose that fight anytime soon. The thing is the fund isn't overweight technology compared to the S&P 500, it's everything else that accounts for the differences in performance. Either way it'll be interesting.
Comments
An interesting Link with some terrific charts. But no great surprise given that the S&P 500 is a capitalization weighted index.
That weighting formula will move any correlation coefficient higher when the overall index is compared to the top 1, or 5, or 10 of its top holdings.
If equally weighted any firm would have a minor impact of that Imdexes movements. However, with the cap weighting operating, the top 5 firms currently make up over 10% of its daily movements. Here is a Link to Morningstar that documents the current weighting distribution for the top current S&P 500 index positions:
http://portfolios.morningstar.com/fund/holdings?t=SPY
I don't know if that weighting has changed dramatically recently, but I'm sure it has indeed changed. That's what makes the marketplace so uncertain and risky.
Best Wishes
What I find interesting is that the FAANG stocks, as a stand alone portfolio, have only a .52 correlation to the market (VOO is almost perfectly correlated at .99). So the FAANG stocks are adding alpha to the S&P 500 index in an uncorrelated way. My question is if FAANG's out performance is uncorrelated to the S&P 500 performance will FAANG's under performance be less of a concern? Said another way, if the FAANG stock perfromance has not caused the rest of the S&P 500 to also "catch fire" why would it's under performance cause the S&P 500 to "crash and burn"?
I believe one could argue that over investment in just a few stocks may suppress investment in the others.
One quick note about DSENX that seemed curious. DSENX has exhibited consistently deeper drawdowns (one could argue only slightly deeper) than VOO (Vanguard S&P 500) shown here (DSENX is in Red):
The derivatives aren't that complicated. If you consider taking a portfolio of $3BN and investing 25% in each of the 4 sectors, and I think it's based on SPDR's Select Sector etfs, then they're doing exactly the same thing. They're just doing it as a "paper bet" with Barclay's where they "settle" the bet at a defined date in the future by the loser paying the winner rather than actually buying and selling all the stocks along the way. Since they still have all the cash in their hands until they have to settle the bet, they invest all the cash in an actively managed collection of bonds. Those bonds offset the cost they pay for making the paper bet, the expense ratio and historically have provided additional income to the fund.
I'm not a tax expert but I think it's interesting that a decent portion of the fund's distributions for the last few years have been long-term gains. My general understanding is that the gains and losses from the swaps should be ordinary income so I would love to understand how they're distributing the long-term gains that they are.
While CAPE has trounced both over its 4.7y life.
Interesting the degree to which RSP deviates from true equal weighting:
http://gi.guggenheiminvestments.com/products/etf/details?productId=92
Its reasonable to think they're investing in value stocks but don't forget their definition is different than what, M* for instance, calls a value stock. CAPE looks at the price now compared to normalized earnings over 10 years.
Here's a link to a white paper about how they construct the index they're following.
bfjlaward.com/pdf/25981/16-33_Bunn_JPM_1017.pdf
I haven't read and thought in depth about every detail but one very important detail of their "relative CAPE ratio" is that it looks at a sectors CAPE ratio relative to what has been normal for that sector. If I understand correctly, then technology, which might have had a higher CAPE ratio historically because of higher growth, can still be selected when it's current CAPE ratio is higher than other sectors but not high compared to the history for that sector. The discussion starts on page 19 of the paper (recognizing that for some reason the paper starts on page 16).
EDIT: I had posted a comment about the impact of market cap that I've thought more about and decided was incorrect so I've deleted it and added a bit at the end.
If I have this right then whether you end up investing in value or growth, not just because of the difference between current CAPE and current P/E, depends on whether the stocks with the highest market cap are "relative CAPE" based cheap or expensive compared to their own sector. That makes expectations for a "value" investment more questionable and could explain some of the performance history. I don't know the history well enough to make any judgment and apparently that history isn't easy to find. I went back through Doubleline's historical SEC filings and they only started disclosing (in their SEC filings) the sectors that the fund was invested in a little more than a year ago. Technology was in the fund as of March 2016 as well. Maybe it would be possible to find historical information based on Barclay's index but I don't have time at the moment to search for it.
One data point that might not be good enough for drawing conclusions but provides a little perspective is the current investments. The traditional CAPE ratio says Real Estate is the most overvalued now, then Technology, Communication Services (which I think Barclays combines with Technology), Healthcare, Basic Materials, Consumer Cyclical, Utilities, Industrials, Consumer Defensive, Financial Services and Energy. The fund takes the 5 least expensive and picks the 4 with the best momentum, so tech, healthcare and consumer cyclical wouldn't even make it into the momentum analysis based on traditional CAPE but they all make it based on "relative CAPE".
Your speculations about what it's doing make sense. I'm in it and obviously pleased so far but it's one I'd probably sell rather than add to (unlike say, TDVFX) if it starts underperforming.
Are you in it?
and redecides every month, right
TDVFX is a fund I like a lot and I'm not in it yet but I will be eventually even if I have to open an account directly with them.
@davidrmoran, yes, you're right, but in determining how expensive each sector is they use a modified CAPE ratio that causes different results. If I understand correctly they've taken the traditional CAPE and divided it by the historical average CAPE (I believe over 20 years). That causes sectors like technology and healthcare, which both have high CAPE ratios now relative to the other sectors, to be considered "cheap" because on average their CAPE ratios have been higher than other sectors for a long time. Once they get to their version of least expensive then they indeed take the 4 with the best momentum out of the least expensive 5 as you said.
tnx
Thank you so much. And here I was thinking DSENX was somehow going to protect me in an S&P 500 carnage. I added it in a portfolio I was not supposed to sell out of, but now I know it should be a sell candidate per my ANALysis.
I really feel like an idiot. I should have been monitoring DSENX closely.
Great chart and good discussion here.
Seems like another great chart would be of CAPE for S&P495.
https://www.markettamer.com/blog/the-best-performance-from-a-fang-since-dracula
this from just a hasty googling
Ten days ago it diverged significantly downward from SP500 quality indexes: SPHQ, SCHDH, DVY, NOBL, OUSA, and similar. RPS is in the middle of these since midweek last week.
What does this signify --- that CAPE occasionally does get FANGed?