Investment principles

Thrugelmir

darkvader wrote: »

I'd be happy to get, on average, 9% to 12%, over the next 10 years and more with a slight chance of few years hitting 20%.

I wouldn't bank on achieving those returns. To say that hey are optimistic is an understatement. Companies do go bust. Then the investment is worthless. Higher returns are there to compensate for the risk.

Ryan_Futuristics

masonic wrote: »

For sectors and stocks, yes, I'm aware, but for countries/regions (and using a single metric in isolation in this case) I didn't think so - do you have any references? I'm not aware of anything pre-Faber that advocates simply using single country CAPE data to pick where to invest geographically and that's my main issue with this approach (if you haven't gathered already). It is often advised to compare stock valuations within market sectors to avoid industry-related noise - it seems to me that similar logic should apply to the diverse economies of different countries with their different economic, regulatory and political situations. Comparing Russia and the USA on the basis of CAPE alone is likely to lead to a misreading of the situation on some level.

As far as I know Faber's the first person to do it this way - but then in his literature, he doesn't suggest that this is the best way ... He had to design an ETF based on a simple implementation of the principle

His own advice to investors is to use valuation to adjust your regional allocations - assuming yearly rebalancing, this would keep you buying slightly cheap, and reducing exposure as regions got more expensive

If the principle works at the stock level, it should function similarly at the macro level

I should think that method is very important. If it wasn't anyone could consistently pick a market beating portfolio of stocks using simple screening tools. You seem to be suggesting that valuation metrics are magical prognostic tools and that all other considerations can be dispensed with. I can't believe it is that easy. You might be aware of the review by Aswath Damodaran (downloadable here) in which a number of value investing approaches including that of Graham have been discussed at length. This includes a description of Graham's 10 step screening methodology, which is by no means simple and suggests (to me at least) method is not at all arbitrary. The paper also lists several things that can go wrong with simple earnings based valuations, such as the risk that earnings are set to fall rather than prices rising (or the potential for growth is limited), differences in the way earnings are reported, absence of diversification. Also pertinent is the tendency towards moving up the risk scale when investing in this manner. Several of these points are relevant to a country-based CAPE approach, especially considering such an approach is currently telling you to pile into Russia, Brazil and southern/eastern Europe. A stockpicking value investor might avoid companies where earnings are suspect, where the board of directors is not acting in the interests of investors, or where they might be adversely affected by specific difficulties. They would still have plenty of companies to choose from. There are a much more limited number of countries with an investable stockmarket and available valuation data and Faber's approach seems to ignore such potential concerns. This, along with the total reliance on quite a simple valuation metric, is where most of my doubts about the applicability of the value methodology to the selection of countries in which to invest lie.

Towards the end of the review I linked to above is a quote that sums it up quite nicely: "The bottom line is that beating the market is never easy and anyone who argues otherwise is fighting history and ignoring the evidence. Value investing starts with a solid base in behavioral finance and empirical evidence but to be successful with it, you have to bring something to the table, a competitive edge that cannot be easily found in the market, and be consistent about staying true to your core philosophy."

Well sometimes the simplest of principles work surprisingly well



In this backtest, the only principle at work is holding the 25 cheapest stocks - and this would have worked fairly consistently for 50 years

http://www.millennialinvest.com/blog/2014/10/28/the-contrarian-sociopathic-mindset

In Graham's era, backtesting like this would have been incredibly laborious (if even possible) ... so it may be that Graham's method could have been greatly simplified

Sometimes I do wonder: why don't more people beat the index ... But then as Faber says, when you look at how 99% of people invest, they still don't even look at fundamentals ... Most people are still buying expensive and selling to cut losses

masonic

Ryan_Futuristics wrote: »

If the principle works at the stock level, it should function similarly at the macro level

I might have missed it, but I don't think you've put forward any reason why that should necessarily be the case. The principle certainly can work when picking from a few hundred stocks, but applying it to a much smaller number of global markets is quite a different proposition. What I remain unconvinced of is that the extrapolation is supported outside of the data set used to formulate the strategy.

Sometimes I do wonder: why don't more people beat the index ... But then as Faber says, when you look at how 99% of people invest, they still don't even look at fundamentals ... Most people are still buying expensive and selling to cut losses

Well let's set aside the private investors for a moment and look beyond them to the professionals. According to Trustnet, there are about 36 funds available to UK investors within the IMA North America sector that specifically invest in the US. I'm excluding those with less than a 10 year history.

Over 10 years, the L&G US Index (a tracker) ranks 9th out of 36. That means that it just sneaks in as a top quartile performer. That doesn't even allow for survivorship bias - the current number of funds available that invest in the sector is 70, so the number available in 2004 is likely to have been larger than 36.

So, >75% of US funds underperformed the US tracker. No fund outperformed it by a very significant margin, with the exception of 'Schroder US Mid Cap', which perhaps should have been excluded. I suspect several of the others have a mid cap focus as well.

That brings us to the chart and article you linked to above. This chart has a logarithmic scale, so percentage changes will appear the same. The average volatility is shown not to be much higher, and the blue line does not appear significantly more volatile than the grey line within any region of the graph. From crudely comparing the lines, there looks to only have been one 5 year period where the blue line underperformed, but other than that the outperformance is fairly consistent and steady, so there should be no short-term performance based reason why fund managers would want to avoid this approach.

Had they followed it, the fund managers would have been able to generate an outperformance of about 60% (again approximated from the graph between 2004-2014) over those 10 years. So, why were none of them able to achieve this (best outperformance <30% over 10 years)? The data from that article certainly suggests these guys are all going to be out of a job when the first tracker is launched over here that is programmed with value-based principles.

JudyLemieux

Here are the three major investment principle.
1.) Always Invest with a Margin of Safety
2.) Expect Volatility and Profit from It
3.) Know What Kind of Investor You Are

gadgetmind

masonic wrote: »

Over 10 years, the L&G US Index (a tracker) ranks 9th out of 36. That means that it just sneaks in as a top quartile performer.

Out of interest, what have the fees been on that tracker over that period?

Ryan_Futuristics

masonic wrote: »

I might have missed it, but I don't think you've put forward any reason why that should necessarily be the case. The principle certainly can work when picking from a few hundred stocks, but applying it to a much smaller number of global markets is quite a different proposition. What I remain unconvinced of is that the extrapolation is supported outside of the data set used to formulate the strategy.

Well if you imagine stats on a individual football players - speed, goals scored, successful tackles, etc. being used to value players, macro principles would just involve averaging those stats over a whole team

It wouldn't tell you who's going to win, but the team with twice the averaging goal rate and defence rate would probably be the better bet

You're right there are weaknesses with it - there is no perfect valuation method ... In fact I think the closest we've got is Price/Book

Well let's set aside the private investors for a moment and look beyond them to the professionals. According to Trustnet, there are about 36 funds available to UK investors within the IMA North America sector that specifically invest in the US. I'm excluding those with less than a 10 year history.

Over 10 years, the L&G US Index (a tracker) ranks 9th out of 36. That means that it just sneaks in as a top quartile performer. That doesn't even allow for survivorship bias - the current number of funds available that invest in the sector is 70, so the number available in 2004 is likely to have been larger than 36.

So, >75% of US funds underperformed the US tracker. No fund outperformed it by a very significant margin, with the exception of 'Schroder US Mid Cap', which perhaps should have been excluded. I suspect several of the others have a mid cap focus as well.

That brings us to the chart and article you linked to above. This chart has a logarithmic scale, so percentage changes will appear the same. The average volatility is shown not to be much higher, and the blue line does not appear significantly more volatile than the grey line within any region of the graph. From crudely comparing the lines, there looks to only have been one 5 year period where the blue line underperformed, but other than that the outperformance is fairly consistent and steady, so there should be no short-term performance based reason why fund managers would want to avoid this approach.

Had they followed it, the fund managers would have been able to generate an outperformance of about 60% (again approximated from the graph between 2004-2014) over those 10 years. So, why were none of them able to achieve this (best outperformance <30% over 10 years)? The data from that article certainly suggests these guys are all going to be out of a job when the first tracker is launched over here that is programmed with value-based principles.

Great questions - the main thing with the US markets is they're so heavily analysed and over-invested, it's hard to find any opportunities

Even Warren Buffett (whose average real return over 50 years has been 20%) failed to beat the S&P500 recently

Buffett's bet that no one could pick a hedge fund that would beat it now comes down to how well he knows it ... and I'm not aware of any UK brokers recommending US active funds

(The other factor is that for 5 years US markets have been rising on stimulus - active managers have been too defensively positioned because markets shouldn't be rising, but these are very artificial circumstances)


Re: why isn't there a US fund that simply invests in the worst 25 stocks in the market?

Unfortunately it would be inherently self-defeating ... As soon as it had a run of good performance and attracted investors, valuations would rocket and it'd turn into the worst fund you could hold


This is the crux of active management's problem - the more popular your fund or investment style, the less value there is in it ... People underestimate the skill it takes just to match an index against this effect

But this is also a problem for indexes - as they become more popular, their valuations are pushed up in just the same way ... No one knows when, but at some point, an index will have a valuation drag in the same as any huge hedge fund ... and this is why smart-beta ETFs, etc are being brought in now

masonic

gadgetmind wrote: »

Out of interest, what have the fees been on that tracker over that period?

The unit class used in the comparison was the expensive retail class (R Inc) as this is the only one with 10 years of data. It currently has an OCF of 0.82%. I might be misremembering, but I have a feeling I saw it on HL many years ago with an AMC of a nice round 1%, so it has probably always been on the expensive side. There must have been a bit of tracking error on its side, because it only lagged the index by a couple of percent, which is less than the effect of charges over the period, but the cheaper HSBC American Index was only a few places lower down the table and performance was a bit closer to expectations in that case.

TCA

Ryan_Futuristics wrote: »

Well if you imagine stats on a individual football players - speed, goals scored, successful tackles, etc. being used to value players, macro principles would just involve averaging those stats over a whole team

It wouldn't tell you who's going to win, but the team with twice the averaging goal rate and defence rate would probably be the better bet

Playing devil's advocate and using your analogy, I'd say that some teams might play in different leagues, therefore you're not comparing like with like.......

masonic

Ryan_Futuristics wrote: »

Great questions - the main thing with the US markets is they're so heavily analysed and over-invested, it's hard to find any opportunities

Even Warren Buffett (whose average real return over 50 years has been 20%) failed to beat the S&P500 recently

Buffett's bet that no one could pick a hedge fund that would beat it now comes down to how well he knows it ... and I'm not aware of any UK brokers recommending US active funds

(The other factor is that for 5 years US markets have been rising on stimulus - active managers have been too defensively positioned because markets shouldn't be rising, but these are very artificial circumstances)

I'm sure if you took a look at the article I linked to above, the following graph won't have escaped you:


So it's pretty clear 2007-2011 was a pretty torrid time for value investors, so there is certainly something about the recent climate that has gone against the value investor.

Re: why isn't there a US fund that simply invests in the worst 25 stocks in the market?

Unfortunately it would be inherently self-defeating ... As soon as it had a run of good performance and attracted investors, valuations would rocket and it'd turn into the worst fund you could hold


This is the crux of active management's problem - the more popular your fund or investment style, the less value there is in it ... People underestimate the skill it takes just to match an index against this effect

But this is also a problem for indexes - as they become more popular, their valuations are pushed up in just the same way ... No one knows when, but at some point, an index will have a valuation drag in the same as any huge hedge fund ... and this is why smart-beta ETFs, etc are being brought in now

These are not new - VTV, the Vanguard Value ETF has been available on the NYSE for over 10 years and has been doing quite well, although not quite as well as the 'worst 25 stocks' strategy would have done. I'm not sure how self defeating products like this would actually be. A simple product should be beatable by a skilled value investor, just by improving the methodology a bit to avoid the dead wood. Judging by the complexity of Graham's screening methodology, he certainly seemed to advocate a more holistic appraisal of value stocks. I suppose for the company running the product, being swamped with cash and becoming too big to take advantage of opportunities probably goes in the category of a nice problem to have. An equilibrium will no doubt be reached that is still ahead of a standard market-weight capitalised tracker and they'll deduct their fees regardless. Maybe this is something GVAL will have to contend with one day.

masonic

TCA wrote: »

Playing devil's advocate and using your analogy, I'd say that some teams might play in different leagues, therefore you're not comparing like with like.......

Yes, and globally they'd even be playing with different balls.