The Lemon Fool

Hi All.

There seems to be a bit of a steepening of the US yield curve. Within this, on Friday, the US 10 year closed up 3/10th's of a basis point @ 1.169%, but is up around 8 bp's over a week and 25 bp's over a month.

Yet, on Friday, here was the behaviour of my four relevant holdings (weekly numbers in brackets) ...

IPE: +4.12% (+1.86%)
HDIV: +3.13% (+0.98%)
CMHY: +1.63% (+2.05%)
VAGP: +0.00% (+1.46%) - I note this one is not high yield - for comparison only

Anyone got any opinion on any of Friday's movements, or are these just semi-random?

Regards, Newroad

Something interesting happened towards the end of Friday's trading. Several of my holdings in 'growth' ITs rose remarkably and coincidently against the flow of the general market. These arose towards the close of trading. Noting ATT +1.1%, HGT+3.6%, MNKS+.08% and SMT+2.6%. I suspect these are not random rises. Would be grateful for other views.

Earlier this evening I posted a reply to a thread on the IT board about SMIF, another high yield bond fund, that had a sharp upward spike in price on Friday. It was pointed out that a large trade volume at the end of the day was due to Uncrossing Trades (UT). In the case of IPE, the large volume trade at the end of the day was designated with LRGS (Large in Scale Deferral). This is all a bit too technical for me but the fact that all of these large volumes appear to be pushing up prices does suggest there is a common factor here.
I agree that it's a bit counter-intuitive to see high yield bond ITs rising in price when US Treasuries have been falling in price and the outlook for inflation, in my opinion, looks to be deteriorating (rising inflation).

Y

Thanks, YieldHog.

Whilst there may have been an uptick on HDIV and to a lesser extent, IPE, near or during the closing auction, they were fairly high all Friday, so I don't think that was the reason on this particular occasion. Indeed, earlier in the day, the price was higher for IPE (at least as high as 73.49 from my quick look, closing at 73.30). There were a couple of large trades at 72.50 totalling £100K or so.

I'm wondering if investment grade stuff is getting less attractive, then some might be (temporarily at least) chasing higher up the yield curve. However, that theorising has no basis in any direct knowledge I have.

Regards, Newroad

With government bond yield rising (in UK,US and Europe) all these high yield trust should be falling in value not rising, especially those that hold large quantities of fixed interest prefs and perpetuals and long dated bonds.

It appears that whatever reason one party decided to buy a whole load of these starting about 3 o'clock Friday and to me reflects just how irrational the markets are. If you look at the pattern of the trades and the timing it's pretty clear it's one party only and also against the flow of everyone else and whatever the reason they wanted it done by close of play Friday. They haven't returned today but who's to say what will happen as the day/week progresses

I have the following guess:

The speed that someone wanted to execute the trades and the poor execution they got as a result suggests it was more about getting their position in the right place. Volume priority over price priority. Usually associated with someone on the wrong side of a position, in pain, losing money and closing to resolve it. However, hard to see who would be short but sometimes people with large pockets just get things wrong either through poor judgement or mistakes by junior staff.


All very frustrating as I would like to buy some more of these but feel they are fully priced. I hold IPE.

Gan020 wrote:With government bond yield rising (in UK,US and Europe) all these high yield trust should be falling in value not rising, especially those that hold large quantities of fixed interest prefs and perpetuals and long dated bonds.

It appears that whatever reason one party decided to buy a whole load of these starting about 3 o'clock Friday and to me reflects just how irrational the markets are. If you look at the pattern of the trades and the timing it's pretty clear it's one party only and also against the flow of everyone else and whatever the reason they wanted it done by close of play Friday. They haven't returned today but who's to say what will happen as the day/week progresses

I have the following guess:

The speed that someone wanted to execute the trades and the poor execution they got as a result suggests it was more about getting their position in the right place. Volume priority over price priority. Usually associated with someone on the wrong side of a position, in pain, losing money and closing to resolve it. However, hard to see who would be short but sometimes people with large pockets just get things wrong either through poor judgement or mistakes by junior staff.


All very frustrating as I would like to buy some more of these but feel they are fully priced. I hold IPE.

In the same post you say prices should be falling, and also "hard to see who would be short". Not sure I understand that contradiction.

Hi (Olag?) Gan.

You mention the possibility of these investment trusts holding long dated bonds (and the obvious sensitivity to changes in interest rates). Average duration and/or maturity (either overall or perhaps within grades) is something that they unfortunately seem not to publish in their annual/interim reports. Conversely, this information is freely available for, say, VAGP.

I could probably figure out mean/median maturity to a reasonable degree of accuracy if I could be bothered (but I can't - and this is what I pay the experts to do anyway) though I wouldn't be sure how to cater for perpetual shares and other more exotic variants in such calculations - as well as the possibility that some, in particularly US ones, might be callable.

Regards, Newroad

There has been increasing demand for high yield as confidence rises, just as there has been with equities. Most HY is short dated, so not greatly affected by small upticks in the yield curve. Reduction in credit spreads is dominating here and there is plenty of scope for further reductions. eg The Co-op group 11% I mentioned in another thread is still yielding around 4.5% whilst 5 year gilts are barely positive.

Hi HiRiskPaul.

That rationale makes some sense. Going back to some limited work I did in October*, I calculated correlation to VWRL in particular (repeated below for convenience)

""...
With VWRL (Jun 2019 to Oct 2020 inclusive, VAGP only started then in London)

VAGP: - 0.111


With VWRL (Jun 2012 to Oct 2020 inclusive)

CMHY: 0.797
HDIV: 0.659
IPE: 0.788


With VAGP (Jun 2019 to Oct 2020 inclusive)

CMHY: 0.002
HDIV: 0.253
IPE: -0.238
...

So, as you hypothesise, if rising confidence raises equities, it's likely to have a positive effect on high yield bonds as well.

Regards, Newroad

* https://www.lemonfool.co.uk/viewtopic.php?f=54&t=25981&p=352930&hilit=newroad+correlation#p352930

Newroad wrote:Hi HiRiskPaul.

That rationale makes some sense. Going back to some limited work I did in October*, I calculated correlation to VWRL in particular (repeated below for convenience)

""...
With VWRL (Jun 2019 to Oct 2020 inclusive, VAGP only started then in London)

VAGP: - 0.111


With VWRL (Jun 2012 to Oct 2020 inclusive)

CMHY: 0.797
HDIV: 0.659
IPE: 0.788


With VAGP (Jun 2019 to Oct 2020 inclusive)

CMHY: 0.002
HDIV: 0.253
IPE: -0.238
...

So, as you hypothesise, if rising confidence raises equities, it's likely to have a positive effect on high yield bonds as well.

Regards, Newroad

* https://www.lemonfool.co.uk/viewtopic.php?f=54&t=25981&p=352930&hilit=newroad+correlation#p352930

I missed that thread, otherwise I might have commented at the time. The way to calculate the correlation between 2 assets is to get price snapshots at regular matching intervals, take the natural logarithm of the price movements (eg ln(price[n]/price[n-1]) and then calculate the correlation between the 2 vectors. It is important that the sampling of prices is done at the same time, so closing are good, but not for different markets, and you should ideally correct for dividend payments.

Hi HiRiskPaul.

I'm sure you're right, but too much effort for me (especially the dividend bit) - that's something market data providers can do!

Obviously, what I did isn't great for measuring relative performance, but hopefully it's OK directionally for measuring correlation. However, I might try and do the ln(price[n]/price[n-1] bit and see how much difference it makes.

Regards, Newroad

Newroad wrote:Hi HiRiskPaul.

I'm sure you're right, but too much effort for me (especially the dividend bit) - that's something market data providers can do!

Obviously, what I did isn't great for measuring relative performance, but hopefully it's OK directionally for measuring correlation. However, I might try and do the ln(price[n]/price[n-1] bit and see how much difference it makes.

Regards, Newroad

Correlation tells you nothing about relative performance. An asset rising steadily at 1% per year would be perfectly correlated with one rising steadily at 10% per year.

Agreed HiRiskPaul.

That's what I was alluding to above, though not in as strident terms. Anyway, I've done what you said using the ln(current/previous) then Excel's CORREL function as before, but dividends still not considered. Here are the numbers, new ones in brackets and bold)

With VWRL (Jun 2019 to Oct 2020 inclusive, VAGP only started then in London)

VAGP: - 0.111 (-0.080)

With VWRL (Jun 2012 to Oct 2020 inclusive)

CMHY: 0.797 (-0.022)
HDIV: 0.659 (0.046)
IPE: 0.788 (0.028)

BNKR: 0.840 (0.007)
MNKS: 0.825 (0.161)
MNP: 0.842 (0.013)
SMT: 0.842 (0.074)

With VAGP (Jun 2019 to Oct 2020 inclusive)

CMHY: 0.002 (0.505)
HDIV: 0.253 (0.544)
IPE: -0.238 (0.445)

Not quite what I expected - how about you?

Regards, Newroad

Looks very odd to me. Here are some correlations I just calculated from US listed iShares ETFs:

Correlations
                       S&P        US Treas       Inv grade	
US Treas             -0.39
Inv grade             0.51            0.28	
HY                    0.79           -0.30            0.73

S&P 500          IVV
US Treasuries    GOVT
Inv Grade        USIG
HY               HYG

Based on monthly sampling, total returns Jan 16 to Jan 21.

This is what I would expect to see. Negative correlations between equities and government bonds, then increasing correlation with decreasing credit quality. You are using IT's which may be muddying the water. ITs are actively managed and have varying discount to NAV. Active management itself can drive correlations all over the place if the manager chooses to take risk off the table at the right and/or wrong time.

Investment grade is not normally quite so correlated with equities by the way. This is largely down to the Covid sell off in risk assets last year. Jan 16 to Jan 20, the matrix looks like this:

Correlations
                       S&P        US Treas       Inv grade	
US Treas             -0.31
Inv grade             0.14            0.80	
HY                    0.73           -0.11            0.41

Hi HiRiskPaul.

It's a fairly simply calculation from the existing columns I have - in effect "=ln(a2/a1)" - so I doubt I've got that part wrong. So, perhaps the odd look is a combination of the use of Investment Trusts (with their variance from NAV) and the strange period we've gone through.

I didn't say that Investment Grade was correlated with equities - but rather, High Yield

"... if rising confidence raises equities, it's likely to have a positive effect on high yield bonds as well ..."

unless you're referring to a different comment?

On a related note, the High Yield Investment Trusts once again finished OK today, though with a reversal from earlier in the day (i.e. IPE up early, CMHY and HDIV up late)

CMHY: +2.28%
HDIV: +0.87%
IPE: -0.41%

Regards, Newroad

PS Just noticed this article: "US corporate bonds grow more susceptible to sudden rise in rates - https://www.ft.com/content/85be899b-9b2d-449b-baa6-c93f5e529ec2

Newroad wrote:Hi HiRiskPaul.

It's a fairly simply calculation from the existing columns I have - in effect "=ln(a2/a1)" - so I doubt I've got that part wrong. So, perhaps the odd look is a combination of the use of Investment Trusts (with their variance from NAV) and the strange period we've gone through.

I didn't say that Investment Grade was correlated with equities - but rather, High Yield

"... if rising confidence raises equities, it's likely to have a positive effect on high yield bonds as well ..."

unless you're referring to a different comment?

On a related note, the High Yield Investment Trusts once again finished OK today, though with a reversal from earlier in the day (i.e. IPE up early, CMHY and HDIV up late)

CMHY: +2.28%
HDIV: +0.87%
IPE: -0.41%

Regards, Newroad

PS Just noticed this article: "US corporate bonds grow more susceptible to sudden rise in rates - https://www.ft.com/content/85be899b-9b2d-449b-baa6-c93f5e529ec2

Just to make sure I understand what you are doing, if a1, a2, a3,..aN etc. contain prices for asset A, b1, b2, b3,..bN for asset B. In C2 to C(N) calculate ln(a2/a1), ln(a3/a2),..ln(a(N)/a(N-1)) and in D2 to D(N) calculate ln(b2/b1), ln(b3/b2),..ln(b(N)/b(N-1)). Then CORREL(C2:C(N),D2:D(N)).

Essentially the higher the credit risk of a portfolio of bonds, the higher the correlation with equities and lower the correlation with government bonds. But, as you can see in my comparison between the correlations for 2016/2021 and 2016/2020, when the stock market gets a shock the correlation between equities and investment grade corporate bonds surges upwards. In other words, investment grade does not give anywhere near the protection that might be expected if you just look at correlations in more sedate periods.

Hi HiRiskPaul.

I needed to wait to switch to my home PC to confirm, but I believe so, yes. The first few rows for the VWRL and VAGP correlation (starting with Jul 19 vs Jun 19) were

VWRL Adj. Close: VWRL HRP Calc: VAGP Adj. Close: VAGP HRP Calc

...
67.09: n/a: 24.81: n/a
71.17: 0.059: 25.00: 0.008
69.00: -0.031: 25.58: 0.023
69.86: 0.012: 25.38: -0.008
...

Interestingly, the High Yield Bond Investment Trusts were the worst performers in my portfolio today. I'm glad that I actually only measure twice per month

(1) At the end of month, of my target portfolio - which gives a good approximation of my actual portfolio and is easier to measure - for performance reasons, and
(2) On the Sunday before the 3rd Wednesday of each month - to rebalance as close as practical to my target portfolio - to use II's free investment facility, for new contributions and re-invested dividends as the mechanism

If I didn't, I might spend (even) more time considering these short term moves and perhaps reacting!

Regards, Newroad