"Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

"Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Sun Dec 31, 2017 10:02 am

I recently read the transcript from an episode of Meb Faber's podcast where he discusses his paper written in 2007 that provides a trend following strategy using the S&P 200 day simple moving average to determine buy/sell rules (link below). The strategy originally proposed in the paper has been discussed on this forum before. Historically it has featured higher expected returns, lower volatility, and lower max draw downs. In the podcast, Meb discusses the out of sample results of the strategy over the last 10 years, which have been favorable primarily due to the 2008-2009 financial crisis.

This made me wonder - has anyone studied whether trend following techniques may dampen sequence of returns risk and, thereby, produce higher safe withdrawal rates? I began running some numbers myself and it appeared that the application of a trend following technique may yield higher SWRs and then I stumbled across a recent paper titled "Reducing Sequence Risk Using Trend Following and the CAPE Ratio" by Clare et al (link below). In their paper, the authors conclude:
The risk of experiencing bad investment outcomes at the wrong time, or sequence risk, is a poorly understood, but crucial aspect of the risk faced by investors, in particular those in the decumulation phase of their savings journey, typically over the period of retirement financed by a defined contributions pension scheme. Using US equity return data from 1872-2014 we show how this risk can be significantly reduced by applying trend-following investment strategies. We also demonstrate that knowledge of a valuation ratio such as the CAPE ratio at the beginning of a decumulation period is useful for enhancing sustainable investment income.
All this said, it's difficult to determine whether this strategy will be effective in the future given:
  • Often, strategies lose their effectiveness once observed.
  • Historically, it would have required significantly more effort and cost to employ a trend following strategy, heightening the observed historical premium of the strategy. In the current age of low expense ratios, ETFs, and limited (or no) commissions, what barriers exist? The strategy now just requires a process and the will to follow it.
  • To some, trend following is synonymous with voodoo or tarot cards - that is, there is no rational explanation for it and no expectation that it will persist in the future. The underpinnings of this strategy are rooted in behavioral finance, so what if investors change their behavior?
I've considered using some variant of Meb's trend following strategy for a while, but have stuck to buy-and-hold. This analysis has led me to conclude that it may have value in the decumulation phase to potentially reduce the sequence of return risk.

What do you think?

Transcript of Meb Faber podcast: http://mebfaber.com/2017/12/13/episode- ... llocation/
Link to "Reducing Sequence Risk Using Trend Following and the CAPE Ratio": https://papers.ssrn.com/sol3/papers.cfm ... id=2764933

Random Walker
Posts: 3206
Joined: Fri Feb 23, 2007 8:21 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Random Walker » Sun Dec 31, 2017 11:49 am

I agree that trend following is all behavioral. And I’m becoming more and more convinced that human behavior is very persistent. Time Series Momentum is basically trend following. The strategy buys what has risen in the recent past and sells what has fallen in the recent past. So it is a shifting allocation strategy. There is a great paper from the AQR people that goes over TS MOM very nicely. If you read it, take note of the “smile curve” and the performance of TS Mom in 8 of 10 substantial equity bear markets. TS Mom is uncorrelated to both equities and bonds and can have a significant positive effect on portfolio efficiency. Moreover, momentum can be found in virtually all asset classes and there is a benefit to diversifying a momentum portfolio across asset classes. Momentum in one asset class I believe is uncorrelated to momentum in the others.
I have made an allocation to time series momentum primarily because of its potential to lessen the pain of extended equity bear markets. Thus it can potentially help lessen the sequence of return risk you refer to. Interested to hear what others think of TS Momentum. I’m a huge believer in efficient markets, but it’s hard to ignore the data showing the persistence of human behavior.

Dave

https://www.aqr.com/-/media/files/paper ... esting.pdf

jmk
Posts: 369
Joined: Tue Nov 01, 2011 7:48 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by jmk » Sun Dec 31, 2017 12:09 pm

You can experiment with simple buy-sell rules on Portfoliovisualizer. Not the CAPE ones though. On paper some seem to work.
Last edited by jmk on Sun Dec 31, 2017 5:37 pm, edited 1 time in total.

stlutz
Posts: 4789
Joined: Fri Jan 02, 2009 1:08 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by stlutz » Sun Dec 31, 2017 12:28 pm

It really depends on exactly how you are executing the moving average strategy and what you are comparing it to, and of course the time period being examined.

Obviously, if you're start year is 2006, the moving average strategy significantly reduces downside risk.

On the other hand, using such a strategy in 2010 and 2011 would have netted returns of .27% and -2.49%. Those were years when a traditional 60/40 balanced portfolio would have returned +13.2 and +4.5%. If you retired at the start of 2010, a MA strategy created more sequence risk. (I calculated these using portfolio visualizer).

The thing about moving average systems is that most trades are losers--not that they lose money; rather, most of the time when you sell you end up buying back at a higher price. They do avoid tail risk (Japan '89) quite nicely, but a long grind downward caused by ill-timed trades has a negative impact on portfolio survivability as well.

In terms of actual retirement planning, I tend to be more sympathetic to the approach of creating an income floor with social security, pensions, annuities, and fixed income investments, and then once that it setup you have a risk portion of the portfolio that you can draw down in a variable way. But that's not necessarily the optimal strategy in many situations either.

magneto
Posts: 956
Joined: Sun Dec 19, 2010 10:57 am
Location: On Chesil Beach

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by magneto » Sun Dec 31, 2017 1:58 pm

"Often, strategies lose their effectiveness once observed."
Momentum as far as can tell seems to be unique, in that the wider it is recognised, the more it succeeds (until it doesn't), as noting the presence of momentum in a security or group of securities, ever more climb aboard the bandwagon.
So that might be one worry less.
Reducing 'Sequence of Return Risk' however is a big ask.

"We also demonstrate that knowledge of a valuation ratio such as the CAPE ratio at the beginning of a decumulation period is useful for enhancing sustainable investment income"
This last may be more significant by recognising the shifting balance between upside potential and downside risk over market cycles.
'There is a tide in the affairs of men ...', Brutus (Market Timer)

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Sun Dec 31, 2017 2:34 pm

DallasGuy,

Here is the latest research on reducing "Sequence Risk".
Dr. ERN (Early Retirement Now) has a Phd. in Economics
& has analyzed reducing sequence risk. This series is the
ultimate guide to safe withdrawal rates part 1-22. You
will want to read all the parts & studies.

Enjoy...

https://earlyretirementnow.com/2017/12/ ... nt-timing/

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Sun Dec 31, 2017 3:56 pm

snarlyjack wrote:
Sun Dec 31, 2017 2:34 pm
DallasGuy,

Here is the latest research on reducing "Sequence Risk".
Dr. ERN (Early Retirement Now) has a Phd. in Economics
& has analyzed reducing sequence risk. This series is the
ultimate guide to safe withdrawal rates part 1-22. You
will want to read all the parts & studies.

Enjoy...

https://earlyretirementnow.com/2017/12/ ... nt-timing/
Thanks. I'm aware of ERN's work. I actually dropped him a note a week or so back with a link to the Clare paper. Perhaps we will hear from him on this topic soon enough...
Last edited by Horton on Sun Dec 31, 2017 4:47 pm, edited 1 time in total.

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Sun Dec 31, 2017 4:00 pm

jmk wrote:
Sun Dec 31, 2017 12:09 pm
You can easily experiment with different buy-sell rules on Portfoliovisualizer. On paper some seem to work.
Portfolio Visualizer is a great tool, but the one thing missing from their Moving Averages Model is the ability to model withdrawals. Instead, their Moving Averages Model assumes that you have a lump sum investment, rather than being in the accumulation (dollar cost averaging) or decumulation (sequence risk) phases.

User avatar
JoMoney
Posts: 6149
Joined: Tue Jul 23, 2013 5:31 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by JoMoney » Sun Dec 31, 2017 4:04 pm

"Increase whipsaw risk and lower returns using trend following" by Someguy et al.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Sun Dec 31, 2017 7:10 pm

DallasGuy,

The problem's I have with Meb's Trend following trading strategy is:

1). Trading the market is not that easy (easier said than done).
2). Taxes...If your moving $Millions in & out of the market depending
on a 200ema line your long term/short term taxes could be huge.
3). The new Federal Taxes (and the conservation around it) it seemed
to me the Gov't want's you to be a long term buy & hold investor not
a trader. See capital gains rate vs. dividend rate's.
4). Reducing sequence of risk returns is 1 thing. Trading my portfolio
on a 200 ema line (with taxes involved) is another thing.
5). Long term buy & hold strategy is a winning strategy (see Ronald Read, The Janitor Next Door).

quantAndHold
Posts: 1904
Joined: Thu Sep 17, 2015 10:39 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by quantAndHold » Sun Dec 31, 2017 8:47 pm

There’s some parts of Meb’s scheme that can’t really be modeled in portfolio visualizer. Meb’s strategy reduces whipsawing by only checking the moving average once per month, for example.

I think for someone in the scary “bad sequence of returns can kill me” portion of early retirement, this could possibly be helpful. It’s kind of like buying an insurance policy. You’re trading some upside, and increasing your risk of getting whipsawed, in order to avoid large drawdowns. I give it a qualified maybe.

For people still working and contributing to their retirement accounts, I think buy, hold, and stay the course is still the best action.

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Sun Dec 31, 2017 9:01 pm

quantAndHold wrote:
Sun Dec 31, 2017 8:47 pm
There’s some parts of Meb’s scheme that can’t really be modeled in portfolio visualizer. Meb’s strategy reduces whipsawing by only checking the moving average once per month, for example.
For Portfolio Visualizer, one of the options for moving average timing models for "Trade Execution" is "Trade at end of month price," which is precisely what Meb modeled. His strategy can be modeled perfectly in PV.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by randomguy » Sun Dec 31, 2017 9:07 pm

quantAndHold wrote:
Sun Dec 31, 2017 8:47 pm
There’s some parts of Meb’s scheme that can’t really be modeled in portfolio visualizer. Meb’s strategy reduces whipsawing by only checking the moving average once per month, for example.

I think for someone in the scary “bad sequence of returns can kill me” portion of early retirement, this could possibly be helpful. It’s kind of like buying an insurance policy. You’re trading some upside, and increasing your risk of getting whipsawed, in order to avoid large drawdowns. I give it a qualified maybe.

For people still working and contributing to their retirement accounts, I think buy, hold, and stay the course is still the best action.
They claim higher returns, lower risk and lower draw downs.If that scheme worked, you should be doing it throughout your retirement career.

At some level you have to ask if it even works at reducing risk/draw downs or if you are just curve fitting. It is hard to say. But I can assure you that after the next crash people will be pumping out numerous schemes that would have avoided it. And some of them would even have talked about it ahead of time.:)

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Sun Dec 31, 2017 9:08 pm

snarlyjack wrote:
Sun Dec 31, 2017 7:10 pm
DallasGuy,

The problem's I have with Meb's Trend following trading strategy is:

1). Trading the market is not that easy (easier said than done).
What do you mean? It takes literally five minutes per month to implement Meb's strategy. And if you want, you can buy Cambria's ETF that implements that strategy.
snarlyjack wrote:
Sun Dec 31, 2017 7:10 pm
2). Taxes...If your moving $Millions in & out of the market depending
on a 200ema line your long term/short term taxes could be huge. 3). The new Federal Taxes (and the conservation around it) it seemed
to me the Gov't want's you to be a long term buy & hold investor not
a trader. See capital gains rate vs. dividend rate's.
For taxable accounts, that's a justifiable concern. For many of us, the majority/all of our assets are in tax advantaged accounts, which are fine for Meb's strategy.
snarlyjack wrote:
Sun Dec 31, 2017 7:10 pm
4). Reducing sequence of risk returns is 1 thing. Trading my portfolio
on a 200 ema line (with taxes involved) is another thing.
Why? It's very easy to do and has been practiced by many investors for at least a century.
snarlyjack wrote:
Sun Dec 31, 2017 7:10 pm
5). Long term buy & hold strategy is a winning strategy (see Ronald Read, The Janitor Next Door).
That may or may not be true. Historically, trend following strategies (apart from tax issues) have produced very similar returns to buy-and-hold but with substantially less volatility. What will be best going forward is unknowable.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Sun Dec 31, 2017 9:12 pm

randomguy wrote:
Sun Dec 31, 2017 9:07 pm
quantAndHold wrote:
Sun Dec 31, 2017 8:47 pm
There’s some parts of Meb’s scheme that can’t really be modeled in portfolio visualizer. Meb’s strategy reduces whipsawing by only checking the moving average once per month, for example.

I think for someone in the scary “bad sequence of returns can kill me” portion of early retirement, this could possibly be helpful. It’s kind of like buying an insurance policy. You’re trading some upside, and increasing your risk of getting whipsawed, in order to avoid large drawdowns. I give it a qualified maybe.

For people still working and contributing to their retirement accounts, I think buy, hold, and stay the course is still the best action.
They claim higher returns, lower risk and lower draw downs.If that scheme worked, you should be doing it throughout your retirement career.

At some level you have to ask if it even works at reducing risk/draw downs or if you are just curve fitting. It is hard to say. But I can assure you that after the next crash people will be pumping out numerous schemes that would have avoided it. And some of them would even have talked about it ahead of time.:)
On his podcast, Meb is adamant that the strategy outlined has had very similar returns to buy-and-hold. He says that its primary purpose is a reduction of drawdowns, and historically, it's done that very well.

I don't really think the curve fitting/backtesting argument is very applicable here since the basic notion of trend following in stocks is at least 100 years old. Meb studied the use of different timing periods (shorter and longer periods than 200 days) and found that over the long-term, the results are not impacted very much. The 200 DMA is the most widely used timing metric in the industry and has been for decades. Further, it has had consistent performance across different markets as well, not just the U.S.

Meb says, and I totally agree with him, that the biggest risk of a timing approach like this is that the returns it produces will be different from year to year than that of the broad market. During bull markets, buy-and-hold is likely to outperform timing. For instance, from 2009 until now, buy-and-hold has trounced virtually all 'reasonable' timing approaches. But when you include periods that include bear markets like 2000-2002 and 2008, the value of timing shines through. Interestingly, Meb actually uses a 50/50 approach, where half of his portfolio is buy-and-hold and the other half uses timing, which is also what Paul Merriman does.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Sun Dec 31, 2017 11:04 pm

I' am going to fall back on Jack Bogle's statement
of "I don't know anyone who can time the market or
know anyone who knows anyone who can time the market".

If you have $5,000. in a IRA & want to play around
with timing the market go for it.

I guess I misunderstood the title of this blog. In my mind
reducing sequence risk & market timing are 2 different things.

In my account: My plan in a down market is to average down
& buy more shares at depressed prices.

magneto
Posts: 956
Joined: Sun Dec 19, 2010 10:57 am
Location: On Chesil Beach

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by magneto » Mon Jan 01, 2018 5:38 am

snarlyjack wrote:
Sun Dec 31, 2017 11:04 pm
I' am going to fall back on Jack Bogle's statement
of "I don't know anyone who can time the market or
know anyone who knows anyone who can time the market".
I guess I misunderstood the title of this blog. In my mind
reducing sequence risk & market timing are 2 different things.
In my account: My plan in a down market is to average down
& buy more shares at depressed prices.
When 'averaging down' is the plan to buy at lower than pre-existing book-cost?
If the price even after a significant setback, remains above book-cost, what then?
'There is a tide in the affairs of men ...', Brutus (Market Timer)

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Mon Jan 01, 2018 10:30 am

Magneto,

(this is my last post on this blog).

I average down the old fashioned way.
I average down monthly, it's called DCA
(dollar cost average).

I have $300,000. in a taxable account &
I' am not moving it (for tax reasons, share
position & dividend reasons). I' am into share
position & compounding for the long term not
dancing in the market.

Like I said you can take your $5,000. IRA & do
what ever you want. As for me I' am getting rich!

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Mon Jan 01, 2018 2:07 pm

snarlyjack wrote:
Sun Dec 31, 2017 11:04 pm
I' am going to fall back on Jack Bogle's statement
of "I don't know anyone who can time the market or
know anyone who knows anyone who can time the market".
Most of the time, when people are talking about timing the market and the difficult in doing so, they are referring to its seeming inability to increase returns. Many of the experts in market timing, such as Faber and Merriman, readily admit that long-term returns from market timing will not beat buy-and-hold. But that's not what we're discussing here at all. We're talking about the ability of a rules-based, objective timing approach to reduce drawdowns and sequence of returns risk over a buy-and-hold approach without a one-for-one reduction of returns. The long-term historic data are very clear that this has worked remarkably well.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by randomguy » Mon Jan 01, 2018 2:41 pm

willthrill81 wrote:
Mon Jan 01, 2018 2:07 pm
snarlyjack wrote:
Sun Dec 31, 2017 11:04 pm
I' am going to fall back on Jack Bogle's statement
of "I don't know anyone who can time the market or
know anyone who knows anyone who can time the market".
Most of the time, when people are talking about timing the market and the difficult in doing so, they are referring to its seeming inability to increase returns. Many of the experts in market timing, such as Faber and Merriman, readily admit that long-term returns from market timing will not beat buy-and-hold. But that's not what we're discussing here at all. We're talking about the ability of a rules-based, objective timing approach to reduce drawdowns and sequence of returns risk over a buy-and-hold approach without a one-for-one reduction of returns. The long-term historic data are very clear that this has worked remarkably well.
If you can get market returns with less volatility, you can leverage up the portfolio and get better than market returns with the same volatility. So in a lot of ways we are talking about beating the market.

There have been systems like this since pretty much since the stock markets formed. Some have worked well for long periods of time and then failed. Maybe this will work better but who knows. The one thing that is clearly a killer is inability to stay the course. Switching to this after markets fall 30% and then back to buy and hold after they rise 200% is the way to get slaughtered.

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Mon Jan 01, 2018 2:51 pm

Willthrill81,

Ok...I'll bite.
(Just because today is a holiday & I' am sitting here watching football games).

First of all that isn't what "Sequence of Risk" is about at all.
I suggest you read part 1-23 & really get a handle on sequence of risk.

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

To give you a example: I've been saving money in my portfolio all my life.
I have a $1 Million dollar portfolio. Or, I could have a $10 Million dollar portfolio.
I' am 65 years old now & getting ready to retire (this is when sequence of risk)
starts to show up. Are you seriously suggesting that I start timing the market with
my $10 Million portfolio?

In my situation it would be an extreme taxable event (taxable account). But let's say
it's all in a IRA/401K (best case situation). Imho, it's not a good idea, at all.
Like I said if you have $5,000. & want to play around...go for it. Remember I' am
65 years old now & I want to retire, travel & play with the grandkids. No one
is talking about or thinking about trading the markets off of a 200 ema line.
(I' am being kind & gentle here). But, we love your input at Bogleheads.

stlutz
Posts: 4789
Joined: Fri Jan 02, 2009 1:08 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by stlutz » Mon Jan 01, 2018 3:22 pm

Are you seriously suggesting that I start timing the market with
my $10 Million portfolio?
What are the alternatives you are comparing? If you have $10M, you can do whatever you want and mostly likely be just fine--100%, a 60/40 balanced portfolio, a moving-average trading system, the Larry portfolio--you'd have to go wrong on the consumption side to run out of money.

For the person who has a $1M portfolio, what actual two portfolios are you comparing (1 timing vs 1 allocation-based)?

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Mon Jan 01, 2018 3:36 pm

I don't know what you guy's are talking about?

In my situation I' am 23 years old & have $300,000. in VHDYX.

Let's figure it out:

At age 65 I could very easily have a $10 Million portfolio.
Let's say VHDYX is paying a 3% dividend = $300,000. per
year in qualified dividends with no capital gains.

Let's say I pull out $100,000. to live on. I could still
reinvest $200,000. per year at the low qualified dividend rate.
The portfolio keeps growing...(the beat goes on).

Why in the world would I want to switch out funds & pay a
35% capital gains tax of $3,500,000. that would be crazy &
totality irresponsible.

With $5,000. play around money, it might be fun. But let's
get real here for a moment...

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Mon Jan 01, 2018 9:27 pm

snarlyjack wrote:
Mon Jan 01, 2018 2:51 pm
Willthrill81,

Ok...I'll bite.
(Just because today is a holiday & I' am sitting here watching football games).

First of all that isn't what "Sequence of Risk" is about at all.
I suggest you read part 1-23 & really get a handle on sequence of risk.

https://earlyretirementnow.com/2016/12/ ... t-1-intro/

To give you a example: I've been saving money in my portfolio all my life.
I have a $1 Million dollar portfolio. Or, I could have a $10 Million dollar portfolio.
I' am 65 years old now & getting ready to retire (this is when sequence of risk)
starts to show up. Are you seriously suggesting that I start timing the market with
my $10 Million portfolio?

In my situation it would be an extreme taxable event (taxable account). But let's say
it's all in a IRA/401K (best case situation). Imho, it's not a good idea, at all.
Like I said if you have $5,000. & want to play around...go for it. Remember I' am
65 years old now & I want to retire, travel & play with the grandkids. No one
is talking about or thinking about trading the markets off of a 200 ema line.
(I' am being kind & gentle here). But, we love your input at Bogleheads.
Thanks, but I actually understand sequence of returns risk very well. The primary elements of this risk for retirees are (1) very low returns for the first 10-15 years of retirement and, very closely related, (2) deep drawdowns early in the withdrawal phase that are not recovered fairly quickly. Rules-based market timing cannot really help with the first of those, but historically it has helped significantly with the latter.

Historically, most stock markets around the globe have at some point had 50% or greater drawdowns of 60/40 portfolios. Someone could just as easily say "Are you seriously suggesting that I use a buy-and-hold strategy with zero downside risk protection and potentially experience a 50% drawdown?" There are risks to buy-and-hold and risks to market timing. All investments carry very real risk; investors must determine what mix and magnitude of risks are most concerning for them and the best strategy for them to personally address those risks.

You speak as though implementing a simple, rules-based approach for five minutes once a month is somehow a burden. Personally, I would find having no downside protection at all being a greater burden. But that's just me. I'm not suggesting that you or anyone else change anything that they're doing now or in the future one bit.

Further, I'm not at all saying that buy-and-hold is a bad strategy; indeed, it is a far better strategy than what most investors use. But it is not the only successful strategy, nor do I think that it is universally appropriate for all investors.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Random Walker
Posts: 3206
Joined: Fri Feb 23, 2007 8:21 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Random Walker » Mon Jan 01, 2018 10:42 pm

I too think sequence of returns risk is huge. I think, somewhat ironically, the shorter the time frame, the more important diversification is. Over short time periods, there’s no telling what source of returns will perform best or how any given source will perform at all. Those 5-10 years on either side of retirement are certainly a critical time. I think it makes tremendous sense to diversify at any time, but especially in this period. Diversification starts with the initial stock/bond split. But one can diversify tremendously more. First diversify equities internationally, including emerging markets. Then diversify across the known drivers of equity returns: market, size, value, momentum, profitability. One can then further diversify across other sources of returns that occur across asset classes: Momentum, value, carry, variance risk premium. One can then diversify into other completely independent sources of return: reinsurance, alternative lending.
I think the two tables in Chapter 9 of Larry Swedroe’s factor book are invaluable. The odds of underperformance decrease tremendously over all time frames when one diversifies across factors. One can extrapolate beyond that with the alternatives.
I don’t think most investors appreciate that even for a typical 60/40 portfolio, about 90% of the portfolio risk is attributable to a single factor, market beta.

Dave

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Tue Jan 02, 2018 12:42 pm

The following figure is provided in the article. It provides real Perfect Withdrawal Rates (same as SWRs) for a 20 year period with and without trend following.

Image

I am interested in longer periods during the decumulation phase because retirees are living longer and most of us here are interested in some form of early retirement. So, I modeled a 40 year period with and without trend following using a technique similar to the authors (although I used a 6-month, rather than a 10-month, moving average). I also modeled two different asset allocations - 100% stocks and a 60/40 portfolio rebalanced monthly. For the 60/40 portfolio, only the 60% equity allocation is exposed to trend following. The results are shown below.

Image

Image

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by randomguy » Tue Jan 02, 2018 12:58 pm

willthrill81 wrote:
Mon Jan 01, 2018 9:27 pm


You speak as though implementing a simple, rules-based approach for five minutes once a month is somehow a burden. Personally, I would find having no downside protection at all being a greater burden. But that's just me. I'm not suggesting that you or anyone else change anything that they're doing now or in the future one bit.
In the end the question though is are you actually getting downside protection or not. Because it worked in the past does not imply that it will work in the future. And how much faith do you have in the system. Personally I don't have enough faith to use schemes like this with 100% of my portfolio. I do have enough in things like this and PE10 that I am willing to go from 80/20 to 50/50 when things are looking sketchy.

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Tue Jan 02, 2018 1:12 pm

randomguy wrote:
Tue Jan 02, 2018 12:58 pm
willthrill81 wrote:
Mon Jan 01, 2018 9:27 pm


You speak as though implementing a simple, rules-based approach for five minutes once a month is somehow a burden. Personally, I would find having no downside protection at all being a greater burden. But that's just me. I'm not suggesting that you or anyone else change anything that they're doing now or in the future one bit.
In the end the question though is are you actually getting downside protection or not. Because it worked in the past does not imply that it will work in the future. And how much faith do you have in the system. Personally I don't have enough faith to use schemes like this with 100% of my portfolio. I do have enough in things like this and PE10 that I am willing to go from 80/20 to 50/50 when things are looking sketchy.
It's true that there are no guarantees, but the historic record is very clear that trend following has resulted in far lower drawdowns than buy-and-hold.

What Faber and Merriman both do is invest half of their portfolios in buy-and-hold and the other half in timing strategies. This helps to prevent the "fear of missing out" during bull markets when buy-and-hold often trounces most timing strategies and to prevent the fear of irrecoverable drawdowns in bear markets.

Either way, investors need to be aware of the risks of both buy-and-hold and market timing. I don't think either is appropriate for all investors. I just wish that Bogleheads didn't act as though rules-based market timing is 'voodoo' that strictly reduces risk and returns in direct proportion to the time spent out of the market because the historic data simply do not support that view.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Tue Jan 02, 2018 1:20 pm

DallasGuy wrote:
Tue Jan 02, 2018 12:42 pm
The following figure is provided in the article. It provides real Perfect Withdrawal Rates (same as SWRs) for a 20 year period with and without trend following.

Image

I am interested in longer periods during the decumulation phase because retirees are living longer and most of us here are interested in some form of early retirement. So, I modeled a 40 year period with and without trend following using a technique similar to the authors (although I used a 6-month, rather than a 10-month, moving average). I also modeled two different asset allocations - 100% stocks and a 60/40 portfolio rebalanced monthly. For the 60/40 portfolio, only the 60% equity allocation is exposed to trend following. The results are shown below.

Image

Image
Excellent charts! It's very interesting that when using trend following, the perfect withdrawal rate increased significantly with 100% stocks over the 60/40 portfolio since there has been so little difference between the two when buy-and-hold has been used. I would be interested to see what differences would exist if the stocks were split between U.S. and international, as well as with tilts such as small and value.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
HomerJ
Posts: 11845
Joined: Fri Jun 06, 2008 12:50 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by HomerJ » Tue Jan 02, 2018 3:03 pm

willthrill81 wrote:
Tue Jan 02, 2018 1:12 pm
What Faber and Merriman both do is invest half of their portfolios in buy-and-hold and the other half in timing strategies. This helps to prevent the "fear of missing out" during bull markets when buy-and-hold often trounces most timing strategies and to prevent the fear of irrecoverable drawdowns in bear markets.
So even the experts who devised these systems don't fully trust them.

That's important.

Anyway, keep it simple is my motto... I'm not going go 100% stocks, and use a trend-following algorithm in retirement, especially since the PhD who invested that algorithm only has 50% of his money in it.

I'd rather just have 50% of my money in bonds and CDs which protects me against bear markets, with the other 50% in the stock market. If I need a higher SWR than a 50/50 stocks/bonds portfolio can give, I'll buy a SPIA.
I just wish that Bogleheads didn't act as though rules-based market timing is 'voodoo' that strictly reduces risk and returns in direct proportion to the time spent out of the market because the historic data simply do not support that view.
Every single system that has failed, at one point looked good during back-testing.

I should point out that I think there a lot of ex-post-facto revisions in the academic world. CAPE predictions using data from 1900-1992 look very different from CAPE predictions using data from 1900-2017.

We've had 25 years of CAPE being abnormally high compared to the previous 90 years. Instead of saying CAPE hasn't been predicting very well over the past 25 years, many experts just seem to change the algorithm using the new data, show a curve that matches past history, and says "See it's always worked in the past!".

But in the past, they didn't have the future data to make it fit so well.

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Tue Jan 02, 2018 3:58 pm

willthrill81 wrote:
Tue Jan 02, 2018 1:20 pm
I would be interested to see what differences would exist if the stocks were split between U.S. and international, as well as with tilts such as small and value.
This article may whet your appetite. It's by the same authors and looks at broader asset classes.

lack_ey
Posts: 6630
Joined: Wed Nov 19, 2014 11:55 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by lack_ey » Tue Jan 02, 2018 4:47 pm

Combining time-series momentum (trend) and valuation signals for market timing is nothing new at all, though I guess here the focus is on implications for sequence risk.

Basically, anything that might improve risk/return, especially lowering left tails, is in theory helpful for reducing sequence risk. No big surprise there. The pushback you get is from people skeptical of the improved risk/return.

Thinking a bit, I remember a paper on combining the two concepts for equity market timing. There are others, but... here, "Market Timing: Sin a Little" from Cliff Asness, Antti Ilmanen, and Thomas Maloney in JOIM.

Image

There's also a brief look at timing for bonds with similar concepts.

This all looks kind of weak to me, though potentially mildly useful, if you can actually implement it costlessly. I'm sure it would be hard to stick with, though, and it doesn't inspire lots of confidence.

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Tue Jan 02, 2018 5:28 pm

HomerJ wrote:
Tue Jan 02, 2018 3:03 pm
willthrill81 wrote:
Tue Jan 02, 2018 1:12 pm
What Faber and Merriman both do is invest half of their portfolios in buy-and-hold and the other half in timing strategies. This helps to prevent the "fear of missing out" during bull markets when buy-and-hold often trounces most timing strategies and to prevent the fear of irrecoverable drawdowns in bear markets.
So even the experts who devised these systems don't fully trust them.

That's important.

Anyway, keep it simple is my motto... I'm not going go 100% stocks, and use a trend-following algorithm in retirement, especially since the PhD who invested that algorithm only has 50% of his money in it.

I'd rather just have 50% of my money in bonds and CDs which protects me against bear markets, with the other 50% in the stock market. If I need a higher SWR than a 50/50 stocks/bonds portfolio can give, I'll buy a SPIA.
It's not a matter of 'trusting' the timing system; it's more of a matter of not trusting one's self. Even for many experts who are well versed in history, it can be hard to follow any investment strategy that underperforms the market over a significant period of time. For instance, virtually all timing systems have underperformed buy-and-hold since 2009. That's almost nine years and counting. On a different note, how many people are now saying things like "the value factor is dead" because LCV has fallen behind TSM for the last decade, and even SCV has barely beaten TSM over the same period yet had significantly higher volatility.

Both Faber and Merriman see the value in having a portion of their portfolios protected from the swings of buy-and-hold yet do not want to divest themselves from the returns that buy-and-hold can provide; the fear of missing out is very real for virtually everyone. I see their practice as very similar to tilting one's portfolio in a particular direction without going all-in.

And neither Faber nor Merriman invented timing the market with the 200 DMA. That's been used for about 100 years, perhaps longer. Trend following is hundreds, if not thousands, of years old.
HomerJ wrote:
Tue Jan 02, 2018 3:03 pm
willthrill81 wrote:
Tue Jan 02, 2018 1:12 pm
I just wish that Bogleheads didn't act as though rules-based market timing is 'voodoo' that strictly reduces risk and returns in direct proportion to the time spent out of the market because the historic data simply do not support that view.
Every single system that has failed, at one point looked good during back-testing.

I should point out that I think there a lot of ex-post-facto revisions in the academic world. CAPE predictions using data from 1900-1992 look very different from CAPE predictions using data from 1900-2017.

We've had 25 years of CAPE being abnormally high compared to the previous 90 years. Instead of saying CAPE hasn't been predicting very well over the past 25 years, many experts just seem to change the algorithm using the new data, show a curve that matches past history, and says "See it's always worked in the past!".

But in the past, they didn't have the future data to make it fit so well.
By 'failed', I suppose you mean that it has underperformed buy-and-hold at some point. That's certainly true, and I've not heard anyone here or elsewhere (who isn't selling something) claim otherwise.

If you compare markets over varying geographic areas as well as time, it's clear that CAPE has been a good, though far from perfect, predictor of future market returns. It hasn't been a great predictor of the U.S.'s returns for the last 25 years, but it has been a solid predictor of returns across countries.

That being said, even Faber has said that pricey markets can always get pricier, and cheap markets can always get cheaper.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Tue Jan 02, 2018 5:30 pm

This is the stock series that Taylor has recommended at Bogleheads.

The author John Collins was written a excellent stock market series using Vanguard Funds.
I would suggest everyone read this series. This might help some people see some light at the
end of the tunnel.

Enjoy...

http://jlcollinsnh.com/2012/04/15/stock ... -save-you/

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Tue Jan 02, 2018 6:10 pm

Here is another good article by John Collins.
(This guy has some excellent articles)

Enjoy...The Time Machine...

http://jlcollinsnh.com/2017/07/26/time- ... or-stocks/

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Wed Jan 03, 2018 9:41 am

Here is another installment of charts relevant to the topic at hand.

Yesterday, I provided a couple time series charts that showed how the SWR varied with and without the application of trend following. The charts generally showed that trend following has historically yielded higher SWRs and helped preserve the 4% rule by producing SWRs above 4% for all years.

Below, I have provided a series of charts that provide the balance at the end of the 40 year period with and without trend following. The charts are based on a starting balance of $1M with monthly inflation-adjusted withdrawals equal to 4% of the starting balance. In short, they show the effectiveness of the 4% rule. Similar to yesterday, I have provided the charts for two different asset allocations - one using 100% stocks and the other using 60% stocks and 40% bonds

The first is a time series chart that shows the balance at the end of the 40 year period for each retirement starting date. It notes the number of periods where a buy and hold portfolio would have been fully depleted and also provides the minimum ending trend following balance (which was always observed to be greater than zero).

The second is a scatterplot that shows the relationship between the ending balance with and without trend following. Blue dots below the red equality line indicate that buy and hold wins and blue dots above the red equality line indicate that trend following wins. I noted that trend following wins in most cases and, when it doesn't, the trend following ending balance is often greater than the balance at the beginning of the 40 year period.

100% Stock

Image

Image

60% Stock / 40% Bonds

Image

Image

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Wed Jan 03, 2018 10:28 am

DallasGuy,
(I' am on my way to work).

You be sure to show your graphs to all the dividend investors that
have made $Millions of dollars & are living off of their dividends.
For every graph that you show, I can pick another graph or article to
disprove your theory. Now, I do agree that the 4% guideline (SWR) is
to high (see Early Retire Now/studies). However, a mix of dividend
growth stocks/dividend paying stocks can help the SWR with the dividend yield.

The problem I see with timing the markets are:
1). long/short term capital gains.
2). Whipsaws.
3). Do you really think "my Grandpa" is going to start timing the market?
4). + to many other agruments to even list such as (travel, retirement, watching the news full time, etc.)

If your so interested in "Timing The Markets" why don't you join a market timing blog &
go for it with like minded people. I' am sure Jack Bogle & Vanguard totality
disagree with you (see Warren Buffett $1Million dollar bet with the "hedge Funds"
you can see how that turned out).

I suggest you learn "something new" read this article about Ronald Read.

Enjoy this article...

https://www.cnbc.com/2015/02/09/heres-h ... rtune.html

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Wed Jan 03, 2018 11:45 am

snarlyjack wrote:
Wed Jan 03, 2018 10:28 am
If your so interested in "Timing The Markets" why don't you join a market timing blog &
go for it with like minded people.
snarlyjack -

Perhaps it is worth me taking a step back to explain the purpose behind my posts and the effort associated with them.

First, it is not my intent to convince you or anyone else that this is the only method to invest and plan for retirement. There are many roads that lead to Rome. It sounds like you have an approach that works, so I would encourage you to stick with it.

As I explained in my very first post on this topic, I've considered using some variant of Meb's trend following strategy for a while, but have stuck to buy-and-hold. The purpose of my posts is simply to present original analysis that I have prepared to examine the impact of trend following on SWRs and retirement outcomes (obviously built upon the initial article by Clare et al). I am examining whether I will commit to this strategy and I am doing it in a public forum so that I get feedback.

While I may not have many posts under my belt, I have been on or around this forum for the better part of a decade. I do not believe my recent posts warrant me leaving this forum to find other "like minded people".

User avatar
BlueEars
Posts: 3634
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by BlueEars » Wed Jan 03, 2018 11:50 am

I think there is a place for alternative ideas of investing in this forum. It is best to discuss these things in a collegial fashion with a focus on the methodology. Attacks that border on religious dogma (invoking well known investor names, for example) are off base I think.

So often on this forum market timing discussions devolve into attacks that have little substance and are just upsetting.

User avatar
BlueEars
Posts: 3634
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by BlueEars » Wed Jan 03, 2018 12:05 pm

To the OP: is the CAPE ratio used to modify the trend following or just as a separate variable to consider in the decumulation phase?

Since the CAPE ratio has poor 1 year predictive value, I'd imagine that it doesn't do much for the trend following algorithm employed. Because of huge changes in market dynamics and accounting changes, I personally just look at a 30 year moving window of CAPE.

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Wed Jan 03, 2018 12:26 pm

BlueEars wrote:
Wed Jan 03, 2018 12:05 pm
To the OP: is the CAPE ratio used to modify the trend following or just as a separate variable to consider in the decumulation phase?

Since the CAPE ratio has poor 1 year predictive value, I'd imagine that it doesn't do much for the trend following algorithm employed. Because of huge changes in market dynamics and accounting changes, I personally just look at a 30 year moving window of CAPE.
Good question.

In their paper, Clare et al examine two approaches. The first being a simple trend following strategy using the 200-day moving average. They also explore an additional, more complicated, approach that involves both trend following and the CAPE. The latter seems a bit opaque and "black-boxish" for me. For example, the authors do not seem to explain the explicit mechanics of the strategy. As such, I have largely ignored this portion of the paper and focused on the first approach using Faber's trend following strategy, which is easily to implement* and is described in Faber's 2007 paper.

* In its most simple form, the strategy can be summarized as follows (pg. 21 of the Faber paper):

BUY RULE
Buy when monthly price > 10-month SMA.
SELL RULE
Sell and move to cash when monthly price < 10-month SMA.

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Wed Jan 03, 2018 12:41 pm

DallasGuy,
(I' am at work on a slow snowy day).

Sorry, let's take a step back & think through this...

1st of all I have a B.S. Degree in Finance & have
studied market timing. I have looked at & studied timing
the stock market & forex (currency's) & commodities.

Most of the money managers (fund managers) have a very
hard time even matching the performance of a index fund.
Their resources are much better than ours (retail investors).
This whole idea that market timing is easy is just plain wrong.

The 2nd point I would like to make is market timing off of a
200 ema is very simplified. You can actually buy off the shelf
market timing system's that are way more advanced & complex.
Even with buying the latest greatest system most market timers
can't make money & the majority of them blow up their account's.
If it were easy "everyone" would be doing it successfully. The
thing is they can't...there are very few successful day traders.

The 3rd point is if day traders can't make a profit what makes you
think you can move around a whole portfolio & make a profit.
Remember I' am sitting on $300,000. portfolio do you think I can pay
the short/long term capital gains tax & still make a profit?
I' am very doubtful...

I think it's a nice dream but I don't think it's realistic. If Meb Faber
was so good don't you think he would be retired now sitting
at home trading the markets. You can call me very doubtful.

User avatar
BlueEars
Posts: 3634
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by BlueEars » Wed Jan 03, 2018 12:43 pm

DallasGuy wrote:
Wed Jan 03, 2018 12:26 pm
...
* In it's most simple form, the strategy can be summarized as follows (pg. 21 of the paper):

BUY RULE
Buy when monthly price > 10-month SMA.
SELL RULE
Sell and move to cash when monthly price < 10-month SMA.
In order to avoid some whipsaws, I would favor some pruning rules. That is, sell only if certain fundamentals are not met. Of course, not selling on a signal would mean you are more in the buy-hold camp for awhile. Some pruning rules were discussed here: http://www.philosophicaleconomics.com/2016/01/gtt/ Unfortunately shortly after that presentation by the author, there was a whipsaw in the market as I recall and those particular pruning rules did not save one from a whipsaw. So whipsaws are something to take seriously. Probably best to implement this in a retirement account because of taxes.

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Wed Jan 03, 2018 1:01 pm

BlueEars wrote:
Wed Jan 03, 2018 12:43 pm
DallasGuy wrote:
Wed Jan 03, 2018 12:26 pm
...
* In it's most simple form, the strategy can be summarized as follows (pg. 21 of the paper):

BUY RULE
Buy when monthly price > 10-month SMA.
SELL RULE
Sell and move to cash when monthly price < 10-month SMA.
In order to avoid some whipsaws, I would favor some pruning rules. That is, sell only if certain fundamentals are not met. Of course, not selling on a signal would mean you are more in the buy-hold camp for awhile. Some pruning rules were discussed here: http://www.philosophicaleconomics.com/2016/01/gtt/ Unfortunately shortly after that presentation by the author, there was a whipsaw in the market as I recall and those particular pruning rules did not save one from a whipsaw. So whipsaws are something to take seriously. Probably best to implement this in a retirement account because of taxes.
Thanks for sharing this article. Looks like a good read! And, I definitely agree that you would only want to implement a timing strategy in a retirement account.

lack_ey
Posts: 6630
Joined: Wed Nov 19, 2014 11:55 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by lack_ey » Wed Jan 03, 2018 1:18 pm

snarlyjack wrote:
Wed Jan 03, 2018 12:41 pm
DallasGuy,
(I' am at work on a slow snowy day).

Sorry, let's take a step back & think through this...

1st of all I have a B.S. Degree in Finance & have
studied market timing. I have looked at & studied timing
the stock market & forex (currency's) & commodities.

Most of the money managers (fund managers) have a very
hard time even matching the performance of a index fund.
Their resources are much better than ours (retail investors).
This whole idea that market timing is easy is just plain wrong.

The 2nd point I would like to make is market timing off of a
200 ema is very simplified. You can actually buy off the shelf
market timing system's that are way more advanced & complex.
Even with buying the latest greatest system most market timers
can't make money & the majority of them blow up their account's.
If it were easy "everyone" would be doing it successfully. The
thing is they can't...there are very few successful day traders.

The 3rd point is if day traders can't make a profit what makes you
think you can move around a whole portfolio & make a profit.
Remember I' am sitting on $300,000. portfolio do you think I can pay
the short/long term capital gains tax & still make a profit?
I' am very doubtful...

I think it's a nice dream but I don't think it's realistic. If Meb Faber
was so good don't you think he would be retired now sitting
at home trading the markets. You can call me very doubtful.
First of all, there's a difference between short-term timing (day trading) and medium-term timing, making moves on the scale of months and years (it's more a continuum, but I think this distinction is still useful). We're talking about the latter. Short-term timing incurs higher transaction costs. We all know there's an army of quants and other traders trying to make money from short-term patterns.

If it works, short-term timing is better because you get more opportunities to effectively roll the dice. Even if you're only right 55% of the time, if you place a bet every day, you will be rich quick. That is, as long as there's sufficient average return on each trade relative to the risk—if you're generating a return of 0.02% +/- 10%, then maybe not so rich. On the other hand, if you're right even 60% of the time, if you're only taking a bet every year, there's a good chance you'll fall behind. And a lot of institutions don't have the patience for those kinds of results; there's career risk to consider. For these kinds of reasons, it seems like it would be harder to arbitrage (using the looser meaning, not in the academic risk-free sense) longer-term trades.

Also, consider that short-term timing operates on timescales where the world generally hasn't changed very much. If you look on the order of say 3 years, you may see different investor preferences, discount rates, economic fundamentals, etc. An advantage or information built up over the level of months or years may result in a profitable trade on average that survives transaction costs, whereas if you're chunking this into trades on the order of days, maybe you lose everything and more to transaction costs and the noise.

Nobody has been talking about short-term timing. Nobody is saying Meb Faber has some short-term timing system that will generate free money for him. The kind of braindead 200-day averages or other kind of medium-term trend following is not short-term profitable in a consistent way, and even historically only gives a modest edge. The timescales don't support retirement on the basis of something like that. In many years, you would have the exact allocation as buy-and-hold, and in some other years you would be a relative loser from whipsawing or otherwise trends reversing and being behind.

All that said, you can call me a little skeptical too, and I don't think it would be worth tax costs if paying capital gains tax on transactions. I think our baseline is assuming a tax-advantaged account.
Last edited by lack_ey on Thu Jan 04, 2018 1:57 pm, edited 1 time in total.

User avatar
BlueEars
Posts: 3634
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by BlueEars » Wed Jan 03, 2018 1:34 pm

DallasGuy wrote:
Wed Jan 03, 2018 1:01 pm
BlueEars wrote:
Wed Jan 03, 2018 12:43 pm
DallasGuy wrote:
Wed Jan 03, 2018 12:26 pm
...
* In it's most simple form, the strategy can be summarized as follows (pg. 21 of the paper):

BUY RULE
Buy when monthly price > 10-month SMA.
SELL RULE
Sell and move to cash when monthly price < 10-month SMA.
In order to avoid some whipsaws, I would favor some pruning rules. That is, sell only if certain fundamentals are not met. Of course, not selling on a signal would mean you are more in the buy-hold camp for awhile. Some pruning rules were discussed here: http://www.philosophicaleconomics.com/2016/01/gtt/ Unfortunately shortly after that presentation by the author, there was a whipsaw in the market as I recall and those particular pruning rules did not save one from a whipsaw. So whipsaws are something to take seriously. Probably best to implement this in a retirement account because of taxes.
Thanks for sharing this article. Looks like a good read! And, I definitely agree that you would only want to implement a timing strategy in a retirement account.
I should have included a snippet of that article instead of just a link. In that Philosophicaleconomics article he writes:
That’s exactly what Growth-Trend Timing does. It takes various combinations of high quality monthly coincident recession signals, and directs the moving average strategy to turn itself off during periods when those signals are unanimously disconfirming recession, i.e., periods where they are all confirming a positive fundamental economic backdrop.

The available monthly signals are:

Real Retail Sales Growth (yoy)
Industrial Production Growth (yoy)
Real S&P 500 EPS Growth (yoy), modeled on a total return basis.
Employment Growth (yoy)
Real Personal Income Growth (yoy)
Housing Start Growth (yoy)

The precise timing criterion for GTT is as follows. Take a reliable monthly growth signal, or better, a collection of reliable monthly growth signals that overlap well to describe the total state of the economy:

If, at the close of the month, the growth signals for the prior month are unanimously positive, then go long or stay long for the next month, and ignore the next step.

If, at the close of the month, the growth signals for the prior month are not unanimously positive, then if price is above the 10 month moving average, then go long or stay long for the next month. If price is below the 10 month moving average, sell or stay out for the next month.

snarlyjack
Posts: 780
Joined: Fri Aug 28, 2015 12:44 pm
Location: Montana

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by snarlyjack » Wed Jan 03, 2018 2:08 pm

I think we are making progress...

First of all a investor would want to do this (market timing)
in a IRA with a limited amount of money. I have always said
if you have $5,000. & want to play around, go for it. Doing it for
the whole portfolio would be very questionable.

The real question is: What's the best way to invest? Their are
a lot of investors here who have $1Million or more. In my mind
this is the question we debate everyday.

What should our AA be, what funds should I invest in & how many
funds do I really need. Should I invest in factors or small value stocks
or dividend paying stocks or international funds? The list goes on.

Market timing is just another segment of the conversation. I might
add just a small segment of the whole conversation.

My own belief, when I was studying investing is to look at the most
successful investors I could find & try to learn from them. That
include's Ronald Read (The Janitor Next Door), Warren Buffett,
Kevin O'Leary, Jack Bogle just to name a few. How are they
investing...they have all made $$$Millions. What is their secrets?

The real question is...what are the investing secret's? And how can
I apply them to my situation. I can only suggest...study the great investors.

jmk
Posts: 369
Joined: Tue Nov 01, 2011 7:48 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by jmk » Thu Jan 04, 2018 1:43 pm

DallasGuy wrote:
Sun Dec 31, 2017 4:00 pm
jmk wrote:
Sun Dec 31, 2017 12:09 pm
You can easily experiment with different buy-sell rules on Portfoliovisualizer. On paper some seem to work.
Portfolio Visualizer is a great tool, but the one thing missing from their Moving Averages Model is the ability to model withdrawals. Instead, their Moving Averages Model assumes that you have a lump sum investment, rather than being in the accumulation (dollar cost averaging) or decumulation (sequence risk) phases.
That's true. The monte carlo allows withdrawals (if you use the advanced tab) of x% of x$ but the Timing models don't. Also, data only goes back to 72.

User avatar
Horton
Posts: 115
Joined: Mon Jan 21, 2008 3:53 pm

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by Horton » Sun Jan 07, 2018 10:04 am

BlueEars wrote:
Wed Jan 03, 2018 12:43 pm
In order to avoid some whipsaws, I would favor some pruning rules. That is, sell only if certain fundamentals are not met. Of course, not selling on a signal would mean you are more in the buy-hold camp for awhile. Some pruning rules were discussed here: http://www.philosophicaleconomics.com/2016/01/gtt/ Unfortunately shortly after that presentation by the author, there was a whipsaw in the market as I recall and those particular pruning rules did not save one from a whipsaw. So whipsaws are something to take seriously. Probably best to implement this in a retirement account because of taxes.
BlueEars - thank you for sharing this post as it eventually lead me to this one. The author points out that you need to account for transaction costs, specifically bid/ask losses, when performing backtests on trend following patterns:
And therefore if we want our backtest of Daily Momentum, or of any strategy that exploits a technical price pattern, to be maximally reliable, we need to commit to applying a slip that matches the actual spread in place at the time. For most of the pre-1990s period, this means a slip of 0.60% or above, applied to each round-trip transaction.
By applying the 0.6% cost, it dramatically lowers the effectiveness of the strategy as seen in the SWRs below.

Image

For reference, here was the original version:
DallasGuy wrote:
Tue Jan 02, 2018 12:42 pm

Image
I suspect that there are additional costs that should be reflected as well (e.g., commissions) that may further bring trend following and buy-and-hold into closer alignment.

Clare et al (the authors of the article referenced in the opening post) opine on this as well:
To add further insight into the possible impact of transactions costs we calculated a “break-even” switching fee, which we define as: the one-way transaction cost that would equate the returns on a Trend Following strategy applied to the S&P500 with those produced by a ‘buy-and hold’ investment in the S&P500, over the full sample period. This break-even value turned out to be 1.35%.
In the end, the authors conclude though that historical costs would not have been this high and they obviously won't be this high in the future, so they believe the effectiveness of the strategy will persist.

I'm not so sure and have decided to stay with buy-and-hold.

Feel free to agree or disagree! :D

User avatar
willthrill81
Posts: 6162
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "Reducing sequence risk using trend following and the CAPE ratio" by Clare et al.

Post by willthrill81 » Sun Jan 07, 2018 11:28 am

In a tax advantaged account, there are no tax implications and usually no transaction costs at all when changing investments.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Post Reply