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.
Support MFO
Donate through PayPal
Mutual Fund Plunges 11% Last Week As Hurricane Hits Florida
You are completely missing the point if you look at it the way that you do. It is a 100% non-correlated asset in a portfolio that does not derive any return from the economy, politics, interest rates, etc. If you had actually looked at the return of the fund since it launched in Dec 2013 you would have seen it had beaten the global stock market over that time period with zero correlation. Sure, there is risk in the portfolio, just like in every other investment I'm sure you have, but the driver of returns are completely different than everything in your portfolio. When your stock portfolio dropped -19% from May 2015 through February 2016 Stone Ridge reinsurance was up over +8% with once again, no correlation. The fee is high certainly, but I have no idea how you can say it has "nothing particularly unique" as it is the most unique mutual fund out there. They negotiate contracts with reinsurance companies to get this exposure. Just like when you get into an auto accident and your premiums go up guess what happens when a major hurricane hits???? Premiums go up too and guess who reaps the benefits of those higher premiums???? That's right, investors in the fund. It is not an investment to day trade (nor do they let you do it anyway) - hence the high minimum so people that actually get it can purchase it, and people who do not, don't get access. It is a buy and hold strategy. There is a reason why quite a few reinsurance companies have been around over 200 years.......
Not sure how this has lost money over the past 3 years as it was up 25% over the past 3 years before IRMA and is still up 12.69% with essentially zero volatility (with the exception of the past 2 weeks). It also goes back to portfolio construction and that is why advisers use it. Take a look at the chart of AIG and Berkshire over that time period and then take a look at SRRIX. Volatility in those stocks is HUGE while SRRIX has next to none. It's about building a diversified portfolio where this is just one component. People pull from their portfolios over time and risk adjusted returns is most important. No client closes their eyes for 10 years and then opens them on the ten year anniversary to see how they did. How you get there day by day is most important. Modern portfolio theory.
When everything (stocks, bonds, REITs, and more) rises in synch in a bull market, then necessarily so.
That would be negatively correlated.
Uncorrelated means, roughly speaking, that one has nothing to do with the other.
That's not to say this "detachment" is necessary to have zero correlation. Consider sine and cosine. Zero correlation, but identical except for phase. Rising together in quadrant IV, falling together in quadrant II; only moving in opposite directions in quadrants I and III. The fact that they're in sync half the time and out of sync half the time doesn't mean they have nothing to do with each other. (One of the limitations of linear correlation - it only takes you so far.)
Ideally what you want are assets which all have a positive tilt but are otherwise negatively (not zero) correlated. Such as (sin(x) + x) and (-sin(x) + x). Both oscillating, but exactly inverted, one zigs when the other zags, so that their oscillations cancel out. At the same time, they are both rising (the +x component).
I thought that the only curve that was "100% non-correlated" to all assets was a constant, i.e. something changing at a constant rate ("essentially zero volatility"). That is, cash or "cash plus", i.e. cash-like but with a positive (or negative) alpha. You've found the holy grail, bet the farm, mortgage your kids, go all in and more.
But that's really not what this is. "Sure, there is risk in the portfolio." While I'm not one to say that risk is the same as volatility, an investment with no volatility (zero std dev) would be riskless. Since there's risk, there's volatility as well.
A problem with "essentially zero volatility" things is that they're very stable (by definition) until they're not. Like land sitting near the San Andreas Fault. Or like a fund that rises smartly and smoothly until it drops 11% or so. That's not to say that long term performance may not be fine, or that it's usually uncorrelated with everything (because its returns are so smooth, like cash). But it's not the holy grail that "zero correlation" and "zero volatility" would imply.
cat bonds are also included in a Pioneer offering HNW. the bonds plunged 16% on Friday, before Irma, as some predictions estimated damages of over $200bln and a wipe out of Fla insurers. on Monday those estimates were paired back to about $50bln, with about $20 bln of insured losses so cat bonds are back into recovery. one needs to note that in the Katrina aftermath they, too, have recovered and rallied.
so yes.. these are uncorrelated risks and returns. but, by its nature, people are happy with the perfect correlation in a bull market, with 'know-it-all comments' all around. efficient portfolio construction is not something everyone does as a hobby in their spare time.
@msf, Yes, a real problem with such terms. If the value of diamonds always fell w/ rising stocks and bonds and land, it too would be correlated. But in journalistic (and financial adviser) idiom, rightly or wrongly, an uncorrelated asset is one that does the opposite of, or at least is reliably different from, the rest. In my experience. Can you think of something zero-correlated in the actual investing world? I like the phrase 'usually uncorrelated'. Somewhere (I can't find it) Freud said 'It takes a high degree of sophistication to believe in coincidence.'
"But in journalistic (and financial adviser) idiom, rightly or wrongly, an uncorrelated asset is one that does the opposite of or at least is reliably different from, the rest. In my experience."
Uncorrelated means unreliably different. What you're describing is negative correlation. My experiences with financial writings (I fastidiously avoid advisers) are different. Excerpts from Vanguard and Schwab papers:
two assets that are “uncorrelated” could be expected to show no systematic, linear relationship between their returns over time. ...
Correlation is a statistical measurement used to convey the strength and direction of a linear relationship between two random variables. In finance, these variables can be anything from an individual security to an entire asset class. Increasingly positive (negative) correlation indicates an increasingly strong (inverse) relationship between the two variables, up to 1 (–1), which indicates a perfectly positive (inverse) relationship.
Diversification offers substantial benefits. As noted above, the diversification benefits associated with combining assets into a portfolio are driven primarily by how closely the returns on those assets move together. When two assets move in perfectly positively correlated (correlation of +1), there are no benefits from combining them in a portfolio. When two equally risky assets are perfectly negatively correlated (correlation of -1), combining the two assets into a portfolio can eliminate all volatility. A correlation of zero means that the two assets are uncorrelated and don’t move together at all.
"Can you think of something zero-correlated in the actual investing world? " I thought I addressed that above. Cash. Backing me up is another Schwab paper.
What is the correlation b/w stocks and bonds, would you state?
Notice the qualifiers in your cites. 'Unreliably.' Which means what? 'could be expected to show no systematic, linear relationship' ... 'measurement used to convey the strength and direction'. The weakness of words.
zero correlation and uncorrelated synonymous, sure. Not worth using the term about in the first place.
I did read your cash mention. That's why I used the phrase 'actual investing world'. No one invests in cash.
No way this is going to be a short post, because people have a mistaken fixation on zero, such as: "I'll wait until my returns get back to zero before selling, that way I won't have lost anything."
Similar fixation with zero with respect to volatility and correlation. Volatility is variation about the mean return, not around zero. So two assets could have perfectly inverse (negative) correlation - as differentiated from being uncorrelated - and still both could always be rising.
Even if one chooses to be lax with wording and use "uncorrelated" to mean negatively correlated, this shows that when everything else is going up, an "uncorrelated" (negatively correlated) asset does not need to be going down. It could just be rising less than its mean (average) Thus, it is not mathematically true that "When everything (stocks, bonds, REITs, and more) rises in synch in a bull market, then [an "uncorrelated" asset] must necessarily [go down]".
Say you have two investments, one which rises 1% a month, give or take volatility, and another which rises 1% a month, give or take its volatility. They could have perfect negative correlation. One might rise (in percent) 1 + sin(month #), while the other might rise 1 - sin(month #). These assets move in perfect (inverse) unison. They both rise every month (since sin(x) ≤1), but when one zigs (rises faster than its mean), the other zags (rises slow than its mean).
Why am I going through all this detail? Because aside from revisiting what I tried to explain before, you asked about stocks and bonds. Bonds, on average, rise in every part of the business cycle. So if one looks only at whether an asset class is simply rising or falling (and not looking at how it's performing relative to its mean performance), one would conclude say that bonds must be closely correlated with cash - they both rise together. But that would be wrong. Cash doesn't correlate well with anything, including bonds.
The graph above is from a good Fidelity page on "How to invest using the business cycle". Stocks and bonds tend to be phase shifted, much like sin and cos, but not as cleanly. So while the correlation coefficient between stocks and bonds tends to be a low (positive) number, it doesn't mean that the performance of stocks and bonds is unrelated. (Nor does it mean that when stocks zig up, bonds will zag down, just that they may gain less than their average.)
Which gets us to Vanguard's precise statement that uncorrelated means that one would expect no systematic linear relationship.
First, linear. As I've been stating, a simple phase shift can produce two curves (say, stocks and bonds) with little linear relationship (like sin and cos). But they're hardly unrelated. Checking for phase shifts, autocorrelation, etc. is one of the first things one does in looking for patterns in time sequence data.
The word you're more concerned about is "expected". A thing with random variables is that the unexpected happens. It must, else the variables are not random. Toss a fair coin enough times and you will come up with ten heads in a row. Toss two coins (to represent stocks and bonds), and sooner or later they'll match 10 times, 20 times, however many times you choose, in a row.
In short, zero correlation does not imply independence, and independence does not mean variables must not show correlation over any arbitrary stretch of time (just that they're expected not to).
"zero correlation and uncorrelated synonymous, sure. Not worth using the term about in the first place. "
I think you missed the point of the bolded sentence, viz. that zero correlated (i.e. uncorrelated) assets don't move together at all. That directly contradicts the claim that an "uncorrelated asset is one that does the opposite".
I agree with you that, unlike the other terms above which have clear mathematical meanings, "reliably" (and hence "unreliably") are weak (vague) words. Outside of this thread, I don't think there's any question that these are extremely fuzzy terms.
"Unreliably" didn't come from a citation. I used the negation ("unreliably") of the word you introduced ("reliably"). Here, I took it on faith that you had some meaning in mind when writing "reliably". I tried to make sense of it, and used it accordingly.
Cash in the "actual investing world". Again, I tried to follow your lead. You suggested looking at how terms are used in journalistic idiom.
We can start with the term "riskless investment". As already discussed, "riskless" entails essentially zero volatility, and cash is the only asset that has that attribute (90 day treasuries, the usual benchmark, would meet the requirements for a MMF). If cash is not an investment, then "riskless investment" must be an oxymoron, albeit one in widespread use.
Moving on to Vanguard, that writes: "Cash investments can lower the overall risk of your portfolio and give you a place to hold money while you wait to invest it." Vanguard seems to be of two minds - cash is an investment awaiting investment. This is pretty typical. While one may not personally think of cash as an investment, and some in the "actual investing world" agree, it seems others don't (hence the ambiguity).
Much has been made of the fact that Schwab Intelligent Portfolios® (robo investments) have a high allocation to cash. Schwab calls this portion of the portfolio a "cash investment".
It defends its use of cash on its FAQ page (see the question "How do you approach cash ..."). There, Schwab says that cash provides stability and liquidity, but also diversification and possible inflation protection. The latter two attributes are usually ascribed to investments.
Finally, TD1 wrote of SRRIX as a "100% non-correlated asset". The meaning in that post was pretty clear and precise, regardless of how sloppily "journalists" might use the term. That's what Maurice was writing about, and what you were responding to when you stated that when everything else was moving up, such a non-correlated asset must necessarily move down (move in the opposite direction).
Likewise, when I wrote "unreliably", I was using it to mean the opposite ("un") of whatever "realiably" meant in the post to which I was responding. Context matters.
About cash, all I was pointing out is that no one invests in cash. Actual investing practice. Some like to quip that cash is what you do not in vest in, by definition. Whatever. Not that everyone and his mother does not write 'cash investments' and 'cash waiting to be invested', true. I sometimes write this stuff for a living, and have typed the phrase 'cash investments' many times. I do try not to write 'riskless investment', and have successfully argued for writing around the wording to emphasize the risklessness and get rid of 'investment'. Just to try and adhere to the lawyers' vetting spirit of investor warning, blah blah. It has worked a few times.
About negative correlation vs noncorrelation, I was commenting in the context of Maurice's 'losing money' phrase. That is, necessarily losing money in comparison. Not a good way to phrase it. though. nor, as you explain, even of thinking about it. Like ads for gold. 'Behavior unlike equities', even in a bull market (they never qualify it that way), as it can go ever so much higher. Etc.
Can you think of assets (not cash) that do not move together at all? Interesting about stocks' / bonds' being phase-shifted.
Finally, did I mention to you long ago the first few months of CD player shipments from Japan (35y ago) which had a function button saying 'random' / 'random play', and the customers returned them to the store after they played the CD tracks 1,3,6,3,3 ?
Comments
Regards,
Ted
I'm terribly embarrassed to admit that there's no way that I could even think about owning this fund. Saved from myself yet again!
When everything (stocks, bonds, REITs, and more) rises in synch in a bull market, then necessarily so.
I find that pretty perplexing.
Just 'diversified' today into DLEUX. (After a big runup, of course.)
Uncorrelated means, roughly speaking, that one has nothing to do with the other.
That's not to say this "detachment" is necessary to have zero correlation. Consider sine and cosine. Zero correlation, but identical except for phase. Rising together in quadrant IV, falling together in quadrant II; only moving in opposite directions in quadrants I and III. The fact that they're in sync half the time and out of sync half the time doesn't mean they have nothing to do with each other. (One of the limitations of linear correlation - it only takes you so far.)
Ideally what you want are assets which all have a positive tilt but are otherwise negatively (not zero) correlated. Such as (sin(x) + x) and (-sin(x) + x). Both oscillating, but exactly inverted, one zigs when the other zags, so that their oscillations cancel out. At the same time, they are both rising (the +x component).
I thought that the only curve that was "100% non-correlated" to all assets was a constant, i.e. something changing at a constant rate ("essentially zero volatility"). That is, cash or "cash plus", i.e. cash-like but with a positive (or negative) alpha. You've found the holy grail, bet the farm, mortgage your kids, go all in and more.
But that's really not what this is. "Sure, there is risk in the portfolio." While I'm not one to say that risk is the same as volatility, an investment with no volatility (zero std dev) would be riskless. Since there's risk, there's volatility as well.
A problem with "essentially zero volatility" things is that they're very stable (by definition) until they're not. Like land sitting near the San Andreas Fault. Or like a fund that rises smartly and smoothly until it drops 11% or so. That's not to say that long term performance may not be fine, or that it's usually uncorrelated with everything (because its returns are so smooth, like cash). But it's not the holy grail that "zero correlation" and "zero volatility" would imply.
so yes.. these are uncorrelated risks and returns. but, by its nature, people are happy with the perfect correlation in a bull market, with 'know-it-all comments' all around. efficient portfolio construction is not something everyone does as a hobby in their spare time.
Yes, a real problem with such terms. If the value of diamonds always fell w/ rising stocks and bonds and land, it too would be correlated. But in journalistic (and financial adviser) idiom, rightly or wrongly, an uncorrelated asset is one that does the opposite of, or at least is reliably different from, the rest. In my experience.
Can you think of something zero-correlated in the actual investing world?
I like the phrase 'usually uncorrelated'.
Somewhere (I can't find it) Freud said 'It takes a high degree of sophistication to believe in coincidence.'
https://www.nytimes.com/2017/09/14/opinion/trump-federal-reserve-markets.html
Uncorrelated means unreliably different. What you're describing is negative correlation. My experiences with financial writings (I fastidiously avoid advisers) are different. Excerpts from Vanguard and Schwab papers:
Vanguard Research, Dynamic correlations: The implications for portfolio construction:
Schwab, Schwab Intelligent Portfolios™ Asset Allocation White Paper
"Can you think of something zero-correlated in the actual investing world? "
I thought I addressed that above. Cash. Backing me up is another Schwab paper.
Schwab Intelligent Portfolios™ The Role of Cash Investments in Asset Allocation:
"Relatively few assets have low or negative correlations with each other. However, as a result of its low volatility, cash is an asset class that tends to have lower correlations relative to all other asset classes."
Notice the qualifiers in your cites. 'Unreliably.' Which means what?
'could be expected to show no systematic, linear relationship' ... 'measurement used to convey the strength and direction'. The weakness of words.
zero correlation and uncorrelated synonymous, sure. Not worth using the term about in the first place.
I did read your cash mention. That's why I used the phrase 'actual investing world'. No one invests in cash.
Violent agreement, again.
What did you think of the NYT piece?
Similar fixation with zero with respect to volatility and correlation. Volatility is variation about the mean return, not around zero. So two assets could have perfectly inverse (negative) correlation - as differentiated from being uncorrelated - and still both could always be rising.
Even if one chooses to be lax with wording and use "uncorrelated" to mean negatively correlated, this shows that when everything else is going up, an "uncorrelated" (negatively correlated) asset does not need to be going down. It could just be rising less than its mean (average) Thus, it is not mathematically true that "When everything (stocks, bonds, REITs, and more) rises in synch in a bull market, then [an "uncorrelated" asset] must necessarily [go down]".
Say you have two investments, one which rises 1% a month, give or take volatility, and another which rises 1% a month, give or take its volatility. They could have perfect negative correlation. One might rise (in percent) 1 + sin(month #), while the other might rise 1 - sin(month #). These assets move in perfect (inverse) unison. They both rise every month (since sin(x) ≤1), but when one zigs (rises faster than its mean), the other zags (rises slow than its mean).
Why am I going through all this detail? Because aside from revisiting what I tried to explain before, you asked about stocks and bonds. Bonds, on average, rise in every part of the business cycle. So if one looks only at whether an asset class is simply rising or falling (and not looking at how it's performing relative to its mean performance), one would conclude say that bonds must be closely correlated with cash - they both rise together. But that would be wrong. Cash doesn't correlate well with anything, including bonds.
The graph above is from a good Fidelity page on "How to invest using the business cycle". Stocks and bonds tend to be phase shifted, much like sin and cos, but not as cleanly. So while the correlation coefficient between stocks and bonds tends to be a low (positive) number, it doesn't mean that the performance of stocks and bonds is unrelated. (Nor does it mean that when stocks zig up, bonds will zag down, just that they may gain less than their average.)
Which gets us to Vanguard's precise statement that uncorrelated means that one would expect no systematic linear relationship.
First, linear. As I've been stating, a simple phase shift can produce two curves (say, stocks and bonds) with little linear relationship (like sin and cos). But they're hardly unrelated. Checking for phase shifts, autocorrelation, etc. is one of the first things one does in looking for patterns in time sequence data.
The word you're more concerned about is "expected". A thing with random variables is that the unexpected happens. It must, else the variables are not random. Toss a fair coin enough times and you will come up with ten heads in a row. Toss two coins (to represent stocks and bonds), and sooner or later they'll match 10 times, 20 times, however many times you choose, in a row.
In short, zero correlation does not imply independence, and independence does not mean variables must not show correlation over any arbitrary stretch of time (just that they're expected not to).
"zero correlation and uncorrelated synonymous, sure. Not worth using the term about in the first place. "
I think you missed the point of the bolded sentence, viz. that zero correlated (i.e. uncorrelated) assets don't move together at all. That directly contradicts the claim that an "uncorrelated asset is one that does the opposite".
I agree with you that, unlike the other terms above which have clear mathematical meanings, "reliably" (and hence "unreliably") are weak (vague) words. Outside of this thread, I don't think there's any question that these are extremely fuzzy terms.
"Unreliably" didn't come from a citation. I used the negation ("unreliably") of the word you introduced ("reliably"). Here, I took it on faith that you had some meaning in mind when writing "reliably". I tried to make sense of it, and used it accordingly.
Cash in the "actual investing world". Again, I tried to follow your lead. You suggested looking at how terms are used in journalistic idiom.
We can start with the term "riskless investment". As already discussed, "riskless" entails essentially zero volatility, and cash is the only asset that has that attribute (90 day treasuries, the usual benchmark, would meet the requirements for a MMF). If cash is not an investment, then "riskless investment" must be an oxymoron, albeit one in widespread use.
Moving on to Vanguard, that writes: "Cash investments can lower the overall risk of your portfolio and give you a place to hold money while you wait to invest it." Vanguard seems to be of two minds - cash is an investment awaiting investment. This is pretty typical. While one may not personally think of cash as an investment, and some in the "actual investing world" agree, it seems others don't (hence the ambiguity).
Much has been made of the fact that Schwab Intelligent Portfolios® (robo investments) have a high allocation to cash. Schwab calls this portion of the portfolio a "cash investment".
It defends its use of cash on its FAQ page (see the question "How do you approach cash ..."). There, Schwab says that cash provides stability and liquidity, but also diversification and possible inflation protection. The latter two attributes are usually ascribed to investments.
Finally, TD1 wrote of SRRIX as a "100% non-correlated asset". The meaning in that post was pretty clear and precise, regardless of how sloppily "journalists" might use the term. That's what Maurice was writing about, and what you were responding to when you stated that when everything else was moving up, such a non-correlated asset must necessarily move down (move in the opposite direction).
Likewise, when I wrote "unreliably", I was using it to mean the opposite ("un") of whatever "realiably" meant in the post to which I was responding. Context matters.
About cash, all I was pointing out is that no one invests in cash. Actual investing practice. Some like to quip that cash is what you do not in vest in, by definition. Whatever. Not that everyone and his mother does not write 'cash investments' and 'cash waiting to be invested', true. I sometimes write this stuff for a living, and have typed the phrase 'cash investments' many times. I do try not to write 'riskless investment', and have successfully argued for writing around the wording to emphasize the risklessness and get rid of 'investment'. Just to try and adhere to the lawyers' vetting spirit of investor warning, blah blah. It has worked a few times.
About negative correlation vs noncorrelation, I was commenting in the context of Maurice's 'losing money' phrase. That is, necessarily losing money in comparison. Not a good way to phrase it. though. nor, as you explain, even of thinking about it. Like ads for gold. 'Behavior unlike equities', even in a bull market (they never qualify it that way), as it can go ever so much higher. Etc.
Can you think of assets (not cash) that do not move together at all? Interesting about stocks' / bonds' being phase-shifted.
Finally, did I mention to you long ago the first few months of CD player shipments from Japan (35y ago) which had a function button saying 'random' / 'random play', and the customers returned them to the store after they played the CD tracks 1,3,6,3,3 ?