HYP1 is 22 - thread discussing income and capital diversification

[BullDog]

Because HYP1 really quite rapidly moved away from that broader level of diversification, and into the very highly focussed situation that we've seen for quite some time now -

https://i.imgur.com/DLlaZiB.png

Cheers,

Itsallaguess

But that's exactly what the studies measure and what the quoted risk reductions are referring to!

I'm not aware of any studies that have looked at the risk of rebalanced/tinkered portfolios.

BoE

Could a reasonable proxy for tinkered portfolios be something along the lines of Merchants Trust or City of London, perhaps?

[1nvest]

But that's exactly what the studies measure and what the quoted risk reductions are referring to!

I'm not aware of any studies that have looked at the risk of rebalanced/tinkered portfolios.

BoE

Could a reasonable proxy for tinkered portfolios be something along the lines of Merchants Trust or City of London, perhaps?

For years I've been tracking/benchmarking HYP1 against FT250 and continue to see similar overall broad rewards, i.e. income equalised, taking the same yearly income as provided by HYP1 out of FT250 total returns. Similar residual capital value rewards however FT250 has had higher volatility. Countering that however is that the FT250 is broader, includes a bunch of Investment Trusts and sub-set HYP's, etc. and is internally rebalanced (by the market). Whilst also significantly reducing single stock concentration risk. The FT250 is 'tinkered' in that periodically firms enter/leave via the top/bottom.

That similar rewards, higher volatility in the FT250 compared to HYP1 goes against traditional measures that tend to indicate not rebalancing tending to lead to similar rewards but with higher volatility. So I suspect any studies may just lead to inconclusive findings, or a conclusion drawn but that was specific to a particular N=1 set or time-period.

[Bubblesofearth]

For years I've been tracking/benchmarking HYP1 against FT250 and continue to see similar overall broad rewards, i.e. income equalised, taking the same yearly income as provided by HYP1 out of FT250 total returns. Similar residual capital value rewards however FT250 has had higher volatility. Countering that however is that the FT250 is broader, includes a bunch of Investment Trusts and sub-set HYP's, etc. and is internally rebalanced (by the market). Whilst also significantly reducing single stock concentration risk. The FT250 is 'tinkered' in that periodically firms enter/leave via the top/bottom.

That similar rewards, higher volatility in the FT250 compared to HYP1 goes against traditional measures that tend to indicate not rebalancing tending to lead to similar rewards but with higher volatility. So I suspect any studies may just lead to inconclusive findings, or a conclusion drawn but that was specific to a particular N=1 set or time-period.

HYP1 was taken entirely from the FTSE100 which should therefore be the benchmark.

It would have been interesting if a HYP (HYP2?) had been constructed entirely from FTSE250 shares. Would it have outperformed the FTSE250?

BoE

[88V8]

For years I've been tracking/benchmarking HYP1 against FT250 and continue to see similar overall broad rewards, i.e. income equalised, taking the same yearly income as provided by HYP1 out of FT250 total returns.So I suspect any studies may just lead to inconclusive findings, or a conclusion drawn but that was specific to a particular N=1 set or time-period.

It would have been interesting if a HYP (HYP2?) had been constructed entirely from FTSE250 shares. Would it have outperformed the FTSE250?

You're back to the time frame.
It might have.
But the outcome of any portfolio is highly dependent on its start time.

V8

[Lootman]

HYP1 was taken entirely from the FTSE100 which should therefore be the benchmark.

To my mind there are two distinct purposes of a benchmark.

The first is to see if your individual efforts to beat that benchmark were successful. In that case you would absolutely choose the FTSE-100 if that was the universe of shares that you were selecting from.

The second is to determine if your asset allocation was optimal. And in that case I would personally choose a broader benchmark. After all, some commonly cited alternative strategies are investing in ITs, ETFs, foreign equities and so on.

If I want to understand your skills at selecting shares then I would compare your returns to the FTSE-100. But if I wanted to know whether your choice of investing in what is just 2% of global market cap made sense, then I would choose a global index.

[1nvest]

HYP1 was taken entirely from the FTSE100 which should therefore be the benchmark.

HYP1 equal weighted, so FTSE100 EQ index as a benchmark?

I've only data for 2010 to 2020 inclusive years for FTSE100 equal weighted index total returns readily to hand, but note that over those years at least the EQ total return index outperformed FTSE100 TR by 2.6% annualised.

Pretty much the FTSE100 has been a dogs dinner since 2000.

If generally the index you choose as a benchmark lags everything else, FTSE EQ, FT250, and even foreign indexes that more broadly aligned with those others when mapped to GB Pounds, then the appropriateness of that benchmark is (highly) questionable.

Even a equal weight index tends to result in distortions in sector weightings. Perhaps a equal weighted sectors with equal weighting of stocks within sectors would be a more appropriate benchmark, but I do not know of such a index. At least not for the UK. For US data you might form a index from individual equal weight sector indicies, with those initially equally weighted, but then that's not the same stock set nor currency.

[1nvest]

Found a reference that indicates 2005 to 2015 FTSE EQ relatively outperformed by, coincidentally or not, 2.6% annualised. So looks like at least for the combined overlapping pairing of those 2005 - 2015 and 2010 - 2020 periods that equal weighted outpaced FTSE100 by 2.6% annualised.

HYP1 total return to recent versus FT All Share total return, which is close to FTSE100, indicates a 3.1% annualised difference. So on a rough measure (November versus calendar years, FTAS instead of FT100) and much of the gap/out-performance does appear to be largely a consequence of equal weighting.

[Lootman]

on a rough measure (November versus calendar years, FTAS instead of FT100) and much of the gap/out-performance does appear to be largely a consequence of equal weighting.

But in that case it is still an open question about whether that equal balancing only needs to be initial (the HYP approach) or whether it needs to be maintained?

[dealtn]

HYP1 was taken entirely from the FTSE100 which should therefore be the benchmark.

HYP1 equal weighted, so FTSE100 EQ index as a benchmark?

Possible, but then one is never balanced but the other is rebalanced every 6 months.

It would be an interesting (but time consuming) exercise to look at what investment in FTSE100 EQ would like look on an unrebalanced basis over different periods and start dates.

The thread over on Investment Strategies, on the back of investment in UK top half dividend yielding stocks rebalanced over time, shows the outperformance of such a strategy (as looked at by the Dimson et al research) was that the success of the approach was due to the rebalancing and capturing of value "outing" not through dividend filtering at selection, for instance.

https://www.lemonfool.co.uk/viewtopic.php?f=8&t=36718

[1nvest]

on a rough measure (November versus calendar years, FTAS instead of FT100) and much of the gap/out-performance does appear to be largely a consequence of equal weighting.

But in that case it is still an open question about whether that equal balancing only needs to be initial (the HYP approach) or whether it needs to be maintained?

IIRC John Mauldin's observation were that non-rebalanced equal weighted yielded the higher overall rewards.

... dig ....

https://www.mauldineconomics.com/frontl ... -mwo040309

... can't see the original charts in that though

Now we come to the interesting part. The next-to-the-top line is the original Dow 30, using a price-weighted index, just like the current Dow 30 uses. The only changes in the next 80 years are companies getting bought or dying. That "Original 30" gives us an annual return of 9.6%. Just 0.7% a year, so you might think, not much difference. But if you start with $100 and compound it for 80 years, that 0.7% becomes a quite large differential. With the Dow 30, your $100 would have grown to $96,993 as of December 2008, but the Original 30 would have grown to $161,603.

And there is an even bigger differential if you simply equal-weight the components rather than use a price-weighting methodology. Your $100 grows at a 10.4% clip and becomes $272,554, or almost three times the actual Dow 30.

[1nvest]

Possible, but then one is never balanced but the other is rebalanced every 6 months.

It would be an interesting (but time consuming) exercise to look at what investment in FTSE100 EQ would like look on an unrebalanced basis over different periods and start dates.

The thread over on Investment Strategies, on the back of investment in UK top half dividend yielding stocks rebalanced over time, shows the outperformance of such a strategy (as looked at by the Dimson et al research) was that the success of the approach was due to the rebalancing and capturing of value "outing" not through dividend filtering at selection, for instance.

https://www.lemonfool.co.uk/viewtopic.php?f=8&t=36718

Consider a set of 20 stocks/sectors, a easier 5% round figure initial equal weighting. Knowing that one will rise to become 50% of the total portfolio value at some later date, is it wise to clip the potential cases, such as those that had doubled to being 10% weighting? Clip your winner(s) to add to others that you hope may do better?

If one stock starts at 5% weight, ends a couple of decades later at 45% weight (again keeping figures simple), then you time-averaged 25% weighting into the stock that likely had the highest total return over those years. That relatively high average weighting into a relatively great performer might more than offset many other laggards relatively poor performance.

IIRC most stocks barely pace cash like rewards, its the right tail small sub-set great cases that tends to uplift the broad 'average'. There's also the small number of left tail bad cases, some even totally failing, but where that left tail is finite, -100% maximum loss, whereas the right tail is infinite (but in reality may be 1000% or 2000% (or maybe more)).

Whether chasing value such as high dividend yield overall pays off (or not) is perhaps more of a coin-flip. High yield = distressed and potentially strongly recovering, or continues to fade potentially into bankruptcy. I haven't read through that thread you linked but guessing that it may indicate that value can be a overall positive biased play. IF that positivity compares equally in reward to run-you-winner(s) then the value style could be safer in respect to having less concentration risk. Overall however I'd guess that concentration risk could yield higher overall rewards in being riskier. But equally if your particular case sees such concentration risk actually hit then that could be very painful. For non rebalanced historical evidence will tend to only reveal those that were lucky, where concentration risk didn't hit and lead to closure/capitulation. Perhaps a portfolio where Enron had risen to be 50% of the portfolio weight, or Lloyds Bank into which Jim and Sue opted to lump all-in at a time when the share price was 20 times higher than nowadays.

And is picking stock wiser than just buying the entire haystack (Index). Yes HYP1 has outrun the FTSE100, but relative out-performance looks much less clear compared to other indicies. Unrewarded (uncompensated) idiosyncratic risk that can be diversified away. Why take unnecessary risk if there is no need to do so.

[moorfield]

Could a reasonable proxy for tinkered portfolios be something along the lines of Merchants Trust or City of London, perhaps?

Well as you know I do like to name check CTY here regularly and pose the question of HYPers:

Why buy a single share, for income, yielding less than CTY? It's yield is what I interpret to be the "high" bar. 4.8% today, higher than ULVR, SHEL, AZN, DGE, BA., SVT, to name a few. Diversified too.

[1nvest]

I've equalised the income in the following, each started with £75K and the same income as provided by HYP1 each year drawn taken from total returns.



Most surprising was just how poorly the US S&P500, total return Pound adjusted compared to others. In total return terms (no income drawn) the S&P500's more recent date value compares to that of the FT250's, but after (HYP1) income was drawn ... quite a difference in final values.

The S&P500 and FCIT both dipped quite heavily in the earlier years (dot com bubble bursting), such that being down whilst also taking a relatively high (HYP1) income, pretty much stalled those compared to others. If the portfolio is down down 33% and S&P500 was in total return after a couple of years with no income having been drawn and you also draw a further 10% as a couple of years income, you've near halved the portfolio value in a couple of years, from where it would struggle to maintain former levels of £££ income.

As a 'tweaked' version, Terry's TJH HYP might be a benchmark, which to recent has lagged HYP1 capital value after HYP1 income having been drawn from its total return. Again TJH HYP did suffer a deeper earlier years dip than HYP1, and that lag has followed through to recent times, just in a less extreme case as the S&P500.

The better/worst cases align with how badly or not they dipped in the earlier/first years.

US SCV by the way is US small cap value stocks. A popular choice with some in the US.
FTAS is the FT All Share, which is mostly FT100 such that the two tend to move/be similar.

[Itsallaguess]

I've already provided you with links that show some 90+% of risk relative to the market is removed by a 15 share portfolio.

How do you square such a small uplift in risk with the huge jump in the return from HYP1 relative to the FTSE100?

Because HYP1 really quite rapidly moved away from that broader level of diversification, and into the very highly focussed situation that we've seen for quite some time now -

https://i.imgur.com/DLlaZiB.png

But that's exactly what the studies measure and what the quoted risk reductions are referring to!

Hang on though - don't we have to be a bit careful here?

Are these 'risk studies' that you're citing actually looking at the risk to someone specifically like Doris?

I only ask because most of these types of papers that I've seen over the years have been compiled on a long-term 'Total Return' basis, which as we know would be completely irrelevant to someone who uses HYP1 to simply throw off cash every year to be used as much-needed income, and where the owner does not consider the capital side of the HYP1 income portfolio at all, and certainly has ZERO inclination to start having to sell down holdings at any point to subsidise that important income...

So could you please tell me if any of these cited studies are specifically covering the ongoing risk to dividend-delivery of a highly-concentrated income-portfolio?

I think this is an important point that would benefit from some clarity please.

Cheers,

Itsallaguess

[pyad]

...I only ask because most of these types of papers that I've seen over the years have been compiled on a long-term 'Total Return' basis, which as we know would be completely irrelevant to someone who uses HYP1 to simply throw off cash every year to be used as much-needed income, and where the owner does not consider the capital side of the HYP1 income portfolio at all, and certainly has ZERO inclination to start having to sell down holdings at any point to subsidise that important income...

So could you please tell me if any of these cited studies are specifically covering the ongoing risk to dividend-delivery of a highly-concentrated income-portfolio?

I think this is an important point that would benefit from some clarity please.

Cheers,

Itsallaguess

That's a bit illogical because the investor you define as one who uses HYP1 purely for income with no consideration at all for the capital, and zero inclination to ever sell off shares to subsidise income, would have no concern over the risks of concentration of that capital by your own definition. Somebody that uninterested in capital probably wouldn't even be aware, or want to be aware, of how it fluctuated.

[Itsallaguess]

...I only ask because most of these types of papers that I've seen over the years have been compiled on a long-term 'Total Return' basis, which as we know would be completely irrelevant to someone who uses HYP1 to simply throw off cash every year to be used as much-needed income, and where the owner does not consider the capital side of the HYP1 income portfolio at all, and certainly has ZERO inclination to start having to sell down holdings at any point to subsidise that important income...

So could you please tell me if any of these cited studies are specifically covering the ongoing risk to dividend-delivery of a highly-concentrated income-portfolio?

That's a bit illogical because the investor you define as one who uses HYP1 purely for income with no consideration at all for the capital, and zero inclination to ever sell off shares to subsidise income, would have no concern over the risks of concentration of that capital by your own definition.

Somebody that uninterested in capital probably wouldn't even be aware, or want to be aware, of how it fluctuated.

I know, but citing a study that *might* be talking about 'Total Return risk' as evidence that Doris need not worry about *Risk to her Income* because of any 'evidence' in it, might not be relevant to the actual risk she should be interested in...

Imagine a situation where all dividends drop to zero, but capital goes on a streak to prove brilliant 'Total Returns' over the coming years...

Brilliant from the standpoint of a TR-based study that *might* show that could happen, but completely useless to someone who simply wants to continue going to their bank machine every month to withdraw her dividend income, and that was the point I was trying to make earlier...

Cheers,

Itsallaguess

[1nvest]

What you are saying is that an investor should accept the risk of a highly concentrated source of income, rather than spread the risk around other shares. You cannot state that the income would have been lower, and indeed, that of HYP1 might fall if RIO reduces its currently very high dividends.

I think with Rio it's when rather than if.

But all this surely misses the point; Doris never looks at her portfolio, and is quite happy that the dividends have continued to roll in.
Therefore it is so far, a success.

V8

How's Doris getting on with migrating holdings into ISA, managing splits/mergers, recording and self-assessment reporting dividends?

Or coping with the likes of 2020 having around only half the inflation adjusted income as in 2019. Oh yes, nearly forgot about ... https://seekingalpha.com/article/842731 ... -investing

"HYPers will suffer during these really bad times from widespread dividend cutting and suspension... total portfolio income will fall and that is unavoidable in the poorest years. The strategy is not for people who cannot live with that possibility".

[Bubblesofearth]

Hang on though - don't we have to be a bit careful here?

Are these 'risk studies' that you're citing actually looking at the risk to someone specifically like Doris?

I only ask because most of these types of papers that I've seen over the years have been compiled on a long-term 'Total Return' basis, which as we know would be completely irrelevant to someone who uses HYP1 to simply throw off cash every year to be used as much-needed income, and where the owner does not consider the capital side of the HYP1 income portfolio at all, and certainly has ZERO inclination to start having to sell down holdings at any point to subsidise that important income...

So could you please tell me if any of these cited studies are specifically covering the ongoing risk to dividend-delivery of a highly-concentrated income-portfolio?

I think this is an important point that would benefit from some clarity please.

Cheers,

Itsallaguess

Of course there are no studies where capital is not considered. I would hazard that you will struggle to find many equity investors outside of the TLF HYP boards that are not concerned about capital. I don't actually believe there are many even on here who would pass a lie detector test responding in that manner.

Doris is a handy pseudonym for a long-term buy and hold investor but it's not realistic take it to extremes. Anyone that really is that indolent but still wanted to invest in equities would be best advised to buy a tracker.

BoE

[88V8]

Doris is a handy pseudonym for a long-term buy and hold investor but it's not realistic take it to extremes. Anyone that really is that indolent but still wanted to invest in equities would be best advised to buy a tracker.

Did trackers exist 22 years ago?

V8

[Itsallaguess]

Hang on though - don't we have to be a bit careful here?

Are these 'risk studies' that you're citing actually looking at the risk to someone specifically like Doris?

I only ask because most of these types of papers that I've seen over the years have been compiled on a long-term 'Total Return' basis, which as we know would be completely irrelevant to someone who uses HYP1 to simply throw off cash every year to be used as much-needed income, and where the owner does not consider the capital side of the HYP1 income portfolio at all, and certainly has ZERO inclination to start having to sell down holdings at any point to subsidise that important income...

So could you please tell me if any of these cited studies are specifically covering the ongoing risk to dividend-delivery of a highly-concentrated income-portfolio?

Of course there are no studies where capital is not considered. I would hazard that you will struggle to find many equity investors outside of the TLF HYP boards that are not concerned about capital. I don't actually believe there are many even on here who would pass a lie detector test responding in that manner.

Doris is a handy pseudonym for a long-term buy and hold investor but it's not realistic take it to extremes. Anyone that really is that indolent but still wanted to invest in equities would be best advised to buy a tracker.

Well for the avoidance of doubt, I personally agree with you on the above points, but it was specifically in the context of this HYP1-related thread that I was wanting to clear up any ambiguity around the relevance of TR-based studies on any perceived claims of 'risk reduction', if those studies weren't actually looking at the situation we're talking about in this particular case, where an investor really does care not one iota about underlying capital or share-values, or has any interest to take any capital returns at all from them, and purely expects that ATM to be functioning every month and simply paying out a stream of dividends...

Cheers,

Itsallaguess