Time to get cashed up?

[spasmodicus]

Hi dspp et al.,

A similar question is to what extent any of you are doing optimal portfolio allocation to try and put yourselves on the efficient frontier ?

Again, I’m just doing it in an ad hoc way

But back to the timing question, the Pentad model that you showed is a lot more sophisticated than running averages on a single index – it uses a multi-channel approach, i.e. 5 channels, hence the name, which a) requires a lot of data input and b) introduces complexity in the way relative weights are given to the channels. The thing that got my attention is that Pentad over 38 years of S&P got a CAGR of 12.22%, whereas LTBH on the S&P500 index achieved 11.67%. OK in 38 years, that amounts to an extra $1.2million on the initial $100k, but hardly a stellar improvement. On average, it did 0.58 trades a year, so not much churn, but it still has to be monitored all the time.

I thought about multi channel too, possibly using other global or UK related indices in conjunction with an index tracker ETF such as MIDD e.g. Baltic Shipping Index type stuff, the possibilities are endless. But before moving to that I’d like to get to the bottom of these moving average “crossing” methodologies. My first thought, from a classical signal processing point of view, is that both the SMA and the EMA are rather poor low pass filters which can be improved upon considerably, if low pass is what you want. The apparent fractal behaviour of stockmarket time series is connected with an amplitude spectral density that increases as 1/f towards lower frequencies, so that low frequency components are even more accentuated by applying a low pass filter.

Hmmm, there’s plenty of food for thought there, but maybe I should look at my asset allocation before worrying about something that is only going to win me a % or 2 a year?
best,
S

[sackofspuds]

But removing some equity risk, and perhaps even removing it so much that it makes it all the way back to cash, sometimes might not entail such a huge element of attempting to 'time the market', and might have a much more basic goal of enabling the investor to just feel more comfortable.

Itsallaguess - around 15% cash, or cash equivalents, and quite comfortable that I may actually be less 'profitable' because of this.....

I think this makes a lot of sense. I used to think that anyone not 100% in equities was too conservative and missing out on gains; a wimp in fact. I'm no longer so arrogant.

[dspp]

Hi dspp et al.,

A similar question is to what extent any of you are doing optimal portfolio allocation to try and put yourselves on the efficient frontier ?

Again, I’m just doing it in an ad hoc way

But back to the timing question, the Pentad model that you showed is a lot more sophisticated than running averages on a single index – it uses a multi-channel approach, i.e. 5 channels, hence the name, which a) requires a lot of data input and b) introduces complexity in the way relative weights are given to the channels. The thing that got my attention is that Pentad over 38 years of S&P got a CAGR of 12.22%, whereas LTBH on the S&P500 index achieved 11.67%. OK in 38 years, that amounts to an extra $1.2million on the initial $100k, but hardly a stellar improvement. On average, it did 0.58 trades a year, so not much churn, but it still has to be monitored all the time.

I thought about multi channel too, possibly using other global or UK related indices in conjunction with an index tracker ETF such as MIDD e.g. Baltic Shipping Index type stuff, the possibilities are endless. But before moving to that I’d like to get to the bottom of these moving average “crossing” methodologies. My first thought, from a classical signal processing point of view, is that both the SMA and the EMA are rather poor low pass filters which can be improved upon considerably, if low pass is what you want. The apparent fractal behaviour of stockmarket time series is connected with an amplitude spectral density that increases as 1/f towards lower frequencies, so that low frequency components are even more accentuated by applying a low pass filter.

Hmmm, there’s plenty of food for thought there, but maybe I should look at my asset allocation before worrying about something that is only going to win me a % or 2 a year?
best,
S

Steady on, you'll be thinking like a control systems engineer if you're not careful. Unfortunately this machine has multiple non-linear evolving feedback loops in it. So a read-across is helpful, but not too literally otherwise you'll be in danger of saying "that's all right in practice, but how does it work in theory".

On which note if you scan through the link sets I gave and follow some of the trails you'll find this:
https://blog.thinknewfound.com/2018/02/ ... -tactical/

Which has this interesting set of observations (amongst many others):

One question we often receive is, “is there one approach that is better than another?” Research over the last decade, however, actually shows that they are highly related approaches.

Bruder, Dao, Richard, and Roncalli (2011) united moving-average-double-crossover strategies and time-series momentum by showing that cross-overs were really just an alternative weighting scheme for returns in time-series momentum.[1] To quote,

“The weighting of each return … forms a triangle, and the biggest weighting is given at the horizon of the smallest moving average. Therefore, depending on the horizon n2 of the shortest moving average, the indicator can be focused toward the current trend (if n2 is small) or toward past trends (if n2 is as large as n1/2 for instance).”

Marshall, Nguyen and Visaltanachoti (2012) proved that time-series momentum is related to moving-average-change-in-direction.[2] In fact, time-series momentum signals will not occur until the moving average changes direction. Therefore, signals from a price-minus-moving-average strategy are likely to occur before a change in signal from time-series momentum.

Levine and Pedersen (2015) showed that time-series momentum and moving average cross-overs are highly related.[3] It also found that time-series momentum and moving-average cross-over strategies perform similarly across 58 liquid futures and forward contracts.

Beekhuizen and Hallerbach (2015) also linked moving averages with returns, but further explored trend rules with skip periods and the popular MACD rule.[4] Using the implied link of moving averages and returns, it showed that the MACD is as much trend following as it is mean-reversion.


etc etc. My bold, but I thought the other stuff was worth quoting as well.

The more I think about these things the more I wonder. Let's say I diversified in timing, by using monthly intervals run weekly (so split a portfolio into four chunks) then diversified in strategy (let's say four strategies) that yields 16 portfolios to be managed on a weekly basis. Each probably of four asset classes. Eek, I'd need to automate that. And that is long-only, no leverage. Hmm.

regards, dspp

[Pastcaring]

I never go to cash,simple strategy,watch what the crowd do.Listen to what the experts tell you to do,then do the exact opposite.

Moving averages,head and shoulders patterns,Fibonacci retracement levels,Elliot wave,complicated maths,kondratieff cycles.

You've got to love experts and complicated maths,they have made everybody rich,haven't they?.

I' ve always gone for truth,reality and simplicity ,the things that people avoid.I've found the buy after a 30% ish drop has worked well for me for well over three decades,as any long term chart shows.

Real time,real life is as I showed recently with CBA,see how it goes in real time.

One thing is certain,very very few people will do what I did.Very few people will admit to reality,they seem to convince themselves the more complicated it is the more factual it must be.The chart will never lie.

Always look for the horse in the race called vested interest,it never loses.

Good luck

[scrumpyjack]

Having been virtually 100% in equities for 50 years, and having a deep seated dislike of cash/fixed interest because of the rampant inflation in the '70s especially, I have now moved 15% into cash.

The main factor behind this is that I am retired, so no earnings now or in future and I would rather I had enough in cash for many years rather than be totally exposed if there is a major market collapse both in share prices and dividends. By that I mean what happened in 1974 (70% fall and lots of passed dividends), not the relatively minor blips of 1987 and 2008.

I'm also rebalancing my holdings to have more non UK as I think the UK market will really tank if there is a Corbyn government.

[spasmodicus]

dspp said

Steady on, you'll be thinking like a control systems engineer if you're not careful. Unfortunately this machine has multiple non-linear evolving feedback loops in it. So a read-across is helpful, but not too literally otherwise you'll be in danger of saying "that's all right in practice, but how does it work in theory".

On which note if you scan through the link sets I gave and follow some of the trails you'll find this:
https://blog.thinknewfound.com/2018/02/ ... -tactical/

Which has this interesting set of observations (amongst many others):

well, yeah. I did do a stint as an electronic engineer at one point, before switching to geophysics, both of which involve plenty of smartarse time series analysis. In those fields, things have moved on a bit computation-wise from the sort of things some of these financial guys are discussing. The possible existence of underlying multiple non-linear evolving feedback loops may not be much of a drawback as we aren't trying to do forward modelling, but just trying to say something about the effects of said alleged loops. After all, you don't have to understand DNA to be able to reproduce.


The guys in your link point out

“but trying to diversify your risk over time simply does not work the same way as diversifying your risk across assets”

That’s another way of acknowledging that stockmarket indices are not ergodic and not even stationary. In which case a mickey-mouse running averagey sort of arm wavey method applied in Excel, or however, is never going to cut the mustard. That said, more sophisticated methods are going to have a tough time extracting any kind of useful signal out of the noise. An approach which relies on occasionally making a big bet once or twice a year, on the whole farm in some cases, on a shaky assessment of the real way the market is moving strikes me as pretty risky. A more nuanced approach would be to make smaller bets more often, maybe using the stats to guide a ladder. I'm not convinced that a fully automated system is even possible. The back-test adjust parameters loop, followed by an application phase, followed by a decision to pull out because "the parameters stopped working after 2013", followed by a new back-testing phase is really an admission of failure.

The more I think about these things the more I wonder. Let's say I diversified in timing, by using monthly intervals run weekly (so split a portfolio into four chunks) then diversified in strategy (let's say four strategies) that yields 16 portfolios to be managed on a weekly basis. Each probably of four asset classes. Eek, I'd need to automate that. And that is long-only, no leverage. Hmm.

Blimey, 16 portfolios once a week is 16 times too many. When all is said and done LTBH works OK for about75% of the time and is improved on by the Schiller CAPE strategy of selling everything when CAPE is above 30 and buying back when below 15 (I think those were the numbers for the US market). This means you only have to take action about every 5 years, which means that at my age I probably don’t have many cycles left.

But I'm sure you're right about the diversification. more gold. more junk bonds. more lottery tickets
regards,
S

[Urbandreamer]

well, yeah. I did do a stint as an electronic engineer at one point, before switching to geophysics, both of which involve plenty of smartarse time series analysis. In those fields, things have moved on a bit computation-wise from the sort of things some of these financial guys are discussing. The possible existence of underlying multiple non-linear evolving feedback loops may not be much of a drawback as we aren't trying to do forward modelling, but just trying to say something about the effects of said alleged loops. After all, you don't have to understand DNA to be able to reproduce.

I too know a bit about control systems, having worked in the field some 30 years. It's not so much the multiple feedback loops as the feed forward loops and the fundimental dependence on initial variables (see chaos theory), that get in the way in the short term. Then as you say the "system" that we attempt to model and predict changes and morphs in front of our eyes.

That said, there was a group of achedemics that did produce a working model. I think that they got a Nobel prize for the fact. They set up a fund and it continued to work. For a while, and then came crashing down!

When all is said and done LTBH works OK for about75% of the time and is improved on by the Schiller CAPE strategy of selling everything when CAPE is above 30 and buying back when below 15 (I think those were the numbers for the US market). This means you only have to take action about every 5 years, which means that at my age I probably don’t have many cycles left.

But I'm sure you're right about the diversification. more gold. more junk bonds. more lottery tickets
regards,
S

I'm not convinced that LTBH works ok as much as 75% of the time (assuming individual shares and no sales), but other than that I am in total agreement. IMHO if you find a system/method or idea that beats cash then you have won the game. It really doesn't matter what the system is. Buy an index tracker and benefit from the dividends while it does the roller coaster thing. Buy a bunch of active funds that go out and invest in Amazon or Allibaba. Buy into biotech or life sciences, who cares as long as it works.

I just feel that sometimes (and yes CAPE is one way of spotting the fact, as is coppock) the writing is on the wall and you would be well off considering a change in stratergy. More so when multiple signals start to flag red, or the bell hop (or modern equivelant) gives you a share tip.
I don't tend to use the likes of CAPE and coppock, but there are pleanty of articles at the moment pointing out that more and more of them are showing red, while others are not. My "gut" tells me that something is in the wind.
I'm in the process of considering my stratergy, hence moving part to cash while I consider what to do.

[GeoffF100]

Moving averages,head and shoulders patterns,Fibonacci retracement levels,Elliot wave,complicated maths,kondratieff cycles.

You've got to love experts and complicated maths,they have made everybody rich,haven't they?

The experts with the complicated maths say this is all a load of tosh. There is no convincing evidence that TA fares any better than chance. It is very popular with brokers though.

[GeoffF100]

Here is another proposal:

https://www.vanguardinvestor.co.uk/arti ... value-cape

http://jpm.iijournals.com/content/44/3/43

The Van Guards predict 3-5% annualised stock market returns for the coming decade. There appear to be as many guesses as there are people making them.

[spasmodicus]

Geoff100 made a nice point here

Here is another proposal:

https://www.vanguardinvestor.co.uk/arti ... value-cape

http://jpm.iijournals.com/content/44/3/43

The Van Guards predict 3-5% annualised stock market returns for the coming decade. There appear to be as many guesses as there are people making them.

This particular guess is about as controversial as driving on the left in the UK. Invest in our products, pay the fees/commissions and we will tell you that the returns will be about what the Equity Premium hypothesis suggests that they should be, on average. see
https://en.wikipedia.org/wiki/Equity_premium_puzzle

Well, thanks for that, guys!

The thing is, I wish that I had something that actually qualified as “a strategy”. I’ve lurked on this board and its Motley predecessor for years and tried to absorb advice from erudite and probably, <west midlands accent> considerably richer than me </west midlands accent> posters, but really all I’ve picked up is a lot of sometimes very high minded stuff about HYP investing (it works), TA (it doesn’t work), fundamental analysis (it might work) and financial advisers and fund managers (they are crooks).

Another frequently proffered piece of advice on these boards is “DYOR” (Do Your Own Research), although I am never quite sure whether those that say it mean it, or are covering their asses from possibly being sued if their advice turns out to be wrong, or worse they get flung in jail for being unlicensed to give financial advice. Neither of those outcomes has, to the best of my knowledge, ever happened.

So, I take this thread as a challenge to DMOR and maybe come up with a simple and potentially effective strategy.

regards and good luck to all,
S

[hiriskpaul]

Moving averages,head and shoulders patterns,Fibonacci retracement levels,Elliot wave,complicated maths,kondratieff cycles.

You've got to love experts and complicated maths,they have made everybody rich,haven't they?

The experts with the complicated maths say this is all a load of tosh. There is no convincing evidence that TA fares any better than chance. It is very popular with brokers though.

Different experts say different things. There is a significant body of research that shows time series momentum exists and is exploitable. Here is a recent paper that looks at a very wide range of data:

https://papers.ssrn.com/sol3/papers.cfm ... id=2993026

I do have a number of reservations about using strategies aimed at taking advantage of trend following though, either to enhance returns or to reduce risk, all about whether the anomaly will persist.

Like lots of other market anomalies, such as value and small cap investing, this stuff is widely known now and so may be incorporated into market prices to the extent that any future benefit that can be derived is much reduced compared to what is visible in back tests.

It is much easier to invest in anomalies such as time series momentum now due to widespread availability of ETFs and the low costs at which they can be traded. Also, there has never been a time when it has been so easy for large numbers of investors to gain access to up to date market information at negligible cost.

The underlying source of returns for this anomaly is likely due to behavioural biases of investors, as it is hard to see how it can be explained in terms of increased risk taking. But it is increasingly the case that trading and investing is being done by computers. I think it Is far more likely that computer driven trading and investing will exploit human bahavioural biases rather than reinforce them, so I am concerned that behavioural biases will influence prices far less in future than they have in the past.

[Urbandreamer]

Another frequently proffered piece of advice on these boards is “DYOR” (Do Your Own Research), although I am never quite sure whether those that say it mean it, or are covering their asses from possibly being sued if their advice turns out to be wrong, or worse they get flung in jail for being unlicensed to give financial advice. Neither of those outcomes has, to the best of my knowledge, ever happened.

There are SERIOUS questions about HYP. It has worked well for its intended purpose to date, but those questions exist. I personally suspect that a lump sum, no tinkering, HYP may not do well for periods over 30 years. If it does not then it still may be described as meeting it's original objective, but does that count as "it works"?

re DYOR.

I think that I mentioned in this thread that I needed to invest some JISA money on behalf of my kids. This lunchtime (school holidays) a daughter and I bought some SMT and gave instructions to buy some CTY on the regular trading day. During the process we were asked to confirm that we had read the "KID". Without confirming so I knew that no deal would go through, so we confirmed. I have asked that she read the document using the DYOR abrieviation.

Ok, here is the explanation.
Despite my feelings about the direction of the market I feel that it is worth PAYING to learn, hence the purchase.
She should DYOR to understand why I have put her into those trusts.

In fact the "KID" is useless in providing a reason to choose one or the other, further research would be required. However it is important to understand what you are buying, recognise that it suits your objectives and recognise the risks. The latter is my complaint about the "KID".

In this case she wants to invest in tech stuff (hence SMT), while I think that having something more stolid will help her ride out the associated volitility.
I shall actually put my son into the same trusts but do it on the regular trading day, saving costs at the expense of education. The time difference may mean that he loses rather than gains, but that itself will provide something for them to ponder on.

[GeoffF100]

Geoff100 made a nice point here

Here is another proposal:

https://www.vanguardinvestor.co.uk/arti ... value-cape

http://jpm.iijournals.com/content/44/3/43

The Van Guards predict 3-5% annualised stock market returns for the coming decade. There appear to be as many guesses as there are people making them.

This particular guess is about as controversial as driving on the left in the UK. Invest in our products, pay the fees/commissions and we will tell you that the returns will be about what the Equity Premium hypothesis suggests that they should be, on average. see
https://en.wikipedia.org/wiki/Equity_premium_puzzle

Your link says:

"The process of calculating the equity risk premium, and selection of the data used, is highly subjective to the study in question, but is generally accepted to be in the range of 3–7% in the long-run. Dimson et al. calculated a premium of "around 3–3.5% on a geometric mean basis" for global equity markets during 1900–2005 (2006).[5] However, over any one decade, the premium shows great variability—from over 19% in the 1950s to 0.3% in the 1970s."

The equity risk premium is the extra return over that of bonds, which is currently about 2%. An equity risk premium of 3-7% would make the equity return 5-9%. Vanguard's predicted 3-5% is much smaller than that. They are saying that they expect the additional return from equities over and above that of bonds to be 1-2% over the next ten years. Lean picking, considering the risks involved.

[dspp]

The equity risk premium is the extra return over that of bonds, which is currently about 2%. An equity risk premium of 3-7% would make the equity return 5-9%. Vanguard's predicted 3-5% is much smaller than that. They are saying that they expect the additional return from equities over and above that of bonds to be 1-2% over the next ten years. Lean picking, considering the risks involved.

Loadsa money has been thrown around for the last decade or so via QE. More supply = lower price. So lower bond returns and therefore also lower equity returns.

At the same time as demographics mature globally, and as more countries climb the income ladder, so too are demand pressures easing. Overcapacity, low growth, low profits.

Both explanations are internally consistent, and consistent with each other.

It could indeed be a lean world out there.

regards, dspp

[GeoffF100]

Moving averages,head and shoulders patterns,Fibonacci retracement levels,Elliot wave,complicated maths,kondratieff cycles.

You've got to love experts and complicated maths,they have made everybody rich,haven't they?

The experts with the complicated maths say this is all a load of tosh. There is no convincing evidence that TA fares any better than chance. It is very popular with brokers though.

Different experts say different things. There is a significant body of research that shows time series momentum exists and is exploitable. Here is a recent paper that looks at a very wide range of data:

https://papers.ssrn.com/sol3/papers.cfm ... id=2993026

I do have a number of reservations about using strategies aimed at taking advantage of trend following though, either to enhance returns or to reduce risk, all about whether the anomaly will persist.

Like lots of other market anomalies, such as value and small cap investing, this stuff is widely known now and so may be incorporated into market prices to the extent that any future benefit that can be derived is much reduced compared to what is visible in back tests.

It is much easier to invest in anomalies such as time series momentum now due to widespread availability of ETFs and the low costs at which they can be traded. Also, there has never been a time when it has been so easy for large numbers of investors to gain access to up to date market information at negligible cost.

The underlying source of returns for this anomaly is likely due to behavioural biases of investors, as it is hard to see how it can be explained in terms of increased risk taking. But it is increasingly the case that trading and investing is being done by computers. I think it Is far more likely that computer driven trading and investing will exploit human bahavioural biases rather than reinforce them, so I am concerned that behavioural biases will influence prices far less in future than they have in the past.

Yes, it is well established that a momentum premium exists, and yes moving average trading is a form of moment trading.

Momentum trading does have risks. Momentum trading may well work just because other people are doing it. If the price has moved up, there is the risk that you are paying too much. If the price has moved down, there is the risk that you are selling for too little.

Do the other forms of TA mention work? Probably not.

Behavioural finance is often rather dubious. It is frequently asserted that private investors make less than market returns by buying high and selling low. Certainly, many of them do that, but for every buyer there is a seller. If the private investors are making less than market returns, others must be making more than market returns. It is not clear who that is. Professional investors usually do what their clients demand, so perhaps the answer is other private investors.

[hiriskpaul]

Behavioural finance is often rather dubious. It is frequently asserted that private investors make less than market returns by buying high and selling low. Certainly, many of them do that, but for every buyer there is a seller. If the private investors are making less than market returns, others must be making more than market returns. It is not clear who that is. Professional investors usually do what their clients demand, so perhaps the answer is other private investors.

Never really thought about that, but yes it is curious. I have read that fund managers, brokers and other intermediaries/service providers siphon off something like 0.7% from the value of global stock markets each year in fees. Perhaps a disproportionate amount of that cost hits private investors who trade frequently?

Alternatively, does this private investor behaviour mean that they lose more in taxes than they otherwise would due to more regular taxes on capital gains?

[hiriskpaul]

An overview of the behaviour of private investors:

https://www.umass.edu/preferen/You%20Mu ... 202011.pdf

From this it would seem that there is a net flow from private investors to institutional investors, in addition to losses private investors face through taxes and trading costs.

[GeoffF100]

There is a lot of interesting stuff in that paper.

There is strong evidence that no fund managers can beat the market, except by chance. This is partly because the markets are reasonably efficient, and partly because the fund managers all have the same skills and tools. Most individual investors will be at a disadvantage. The issue is whether the market is inefficient enough for that to matter. The evidence in the paper suggests that it probably is inefficient enough to matter.

I had previously seen studies that found that the stocks that private investors sold did better than those that they bought. If this was a consistent phenomenon, brokers would be able to make a fortune by going long on the stocks that their clients sold and short on those that they bought, with no market risk. I have never heard of any of them doing this. The paper did say that some studies have found that the stocks that individual investors buy outperform in the short term and underperform in the longer term. There are some grounds for scepticism here.

My comment was more addressed to investors who buy funds high and sell them low. Who is the counter party there? Very few funds successfully go into cash at the top of the market and buy at the bottom, but funds that rebalance do that to some degree.

My own behaviour and results are in stark contrast to most of the people in the study. I trade very little. My directly held UK portfolio has very low costs. My alpha over the last three years is 6%, but that is pure luck. I also buy overseas trackers, partly because the costs for direct foreign investment are higher, but also to shelter from CGT.

The best policy here is to assume that you are an idiot, and behave accordingly.

[hiriskpaul]

When funds are discussed, this typically means open ended funds as worldwide they completely dominate compared with closed ended funds. When net cash is withdrawn, this necessitates the fund selling stocks. Those stocks could go to a variety of different types of investors, but it will be mostly institutional buyers, some of which will be other retail funds. Most private investor money sits in funds now.

Regarding brokers attempting to cash in by doing the opposite of their clients, I suspect that profiting from it is beyond the limits of arbitrage. There are also rules brokers need to be careful of breaking. This sort of thing might look like front running.

[Alaric]

There is strong evidence that no fund managers can beat the market, except by chance.

What about beating the market by taking higher risks, a more concentrated portfolio for example? If you define "the market" as the FTSE All share, investing in the top 10 is either going to beat it or under perform it, depending on how the biggest caps perform. It isn't monkeys throwing darts to select the 10 biggest companies, even if you don't know when you do it whether it's a cunning plan or missing out on the growth of smaller companies.