The Ulcer Index

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Kevin K
Posts: 119
Joined: Sun Aug 26, 2007 7:47 pm

The Ulcer Index

Post by Kevin K » Wed Nov 01, 2017 10:41 am

There's already a long thread here on "The Coolest Portfolio Tool on the Web" whose most recent post features Tyler's latest but I thought it was of such interest that it deserved its own.

It seems to me that this new post makes a pretty compelling case for ditching standard deviation and replacing it with "the ulcer index" for asset allocation/portfolio design purposes. Especially for those in retirement (or risk-averse sorts still in the accumulation phase) the difference in the "ride" between, say, the Golden Butterfly or Larry Portfolio and any of the more mainstream stock-and-bond allocations on the site in terms of the real-world return rollercoaster is pretty shocking.

Hats off to Tyler for offering a set of tools that are light years ahead of those offered by any of the large brokerages. I'm kind of amazed Vanguard or DFA hasn't made him a job offer he couldn't refuse.

https://portfoliocharts.com/2017/11/01/ ... more-22337

livesoft
Posts: 69586
Joined: Thu Mar 01, 2007 8:00 pm

Re: The Ulcer Index

Post by livesoft » Wed Nov 01, 2017 12:26 pm

I was just having a discussion on this subject earlier today. I haven't really used the web site. Can one put in a ticker symbol and get an Ulcer Index? For example, VSMGX and VWENX?
Wiki This signature message sponsored by sscritic: Learn to fish.

User avatar
triceratop
Posts: 5838
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: The Ulcer Index

Post by triceratop » Wed Nov 01, 2017 12:48 pm

livesoft wrote:
Wed Nov 01, 2017 12:26 pm
I was just having a discussion on this subject earlier today. I haven't really used the web site. Can one put in a ticker symbol and get an Ulcer Index? For example, VSMGX and VWENX?
Not that I'm aware, but you can simulate the results with an asset allocation, with this tool: https://portfoliocharts.com/portfolio/drawdowns/

The ulcer index for a 36/24/40 TSM/TISM/IT Bond portfolio (close to VSMGX) is 8.9
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."

User avatar
nisiprius
Advisory Board
Posts: 39743
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: The Ulcer Index

Post by nisiprius » Wed Nov 01, 2017 12:51 pm

livesoft wrote:
Wed Nov 01, 2017 12:26 pm
I was just having a discussion on this subject earlier today. I haven't really used the web site. Can one put in a ticker symbol and get an Ulcer Index? For example, VSMGX and VWENX?
I've just spent five minutes clicking around the site and looking for that, but couldn't find anything like that.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

daveydoo
Posts: 1564
Joined: Sun May 15, 2016 1:53 am

Re: The Ulcer Index

Post by daveydoo » Wed Nov 01, 2017 12:51 pm

Kevin K wrote:
Wed Nov 01, 2017 10:41 am

Hats off to Tyler for offering a set of tools that are light years ahead of those offered by any of the large brokerages.
But those charts only include data on the depth and breadth of the drawdowns themselves, without factoring in all the drawdown-free time, right? One big drawdown in an otherwise stellar history -- nothing but gains -- would make an investment look terrible. Not sure this is the one-size-fits-all metric I've been looking for. But I agree that SD is a misleading index. His examples of re-ordering annual-return data to create scenarios of varying "scariness" is also not super illuminating since are we really expecting fifty years where the annual return increases or decreases every single year? These alternate-universe scenarios are mathematical curiosities that would virtually never appear on a Monte Carlo simulation no matter how many times you spun the wheel.
"I mean, it's one banana, Michael...what could it cost? Ten dollars?"

User avatar
nisiprius
Advisory Board
Posts: 39743
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: The Ulcer Index

Post by nisiprius » Wed Nov 01, 2017 12:53 pm

"Ulcers?" Seriously?

I think it's been several decades since people believed that ulcers are caused by the anxiety of having to make difficult executive decisions. Nowadays they seem to think it's Helicobacter pylori.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
Tyler9000
Posts: 479
Joined: Fri Aug 21, 2015 11:57 am

Re: The Ulcer Index

Post by Tyler9000 » Wed Nov 01, 2017 1:20 pm

Thanks Kevin K. I'm glad you found the post useful! I certainly can't take credit for the Ulcer Index idea, but I enjoy being able to spread the word as I do believe it's a helpful way to look at portfolio volatility.

livesoft wrote:
Wed Nov 01, 2017 12:26 pm
I was just having a discussion on this subject earlier today. I haven't really used the web site. Can one put in a ticker symbol and get an Ulcer Index? For example, VSMGX and VWENX?
That level of data integration requires an expensive subscription I can't afford. But as Triceratop illustrates, you can approximate the returns by modeling a portfolio with a similar construction.

daveydoo wrote:
Wed Nov 01, 2017 12:51 pm
But those charts only include data on the depth and breadth of the drawdowns themselves, without factoring in all the drawdown-free time, right? One big drawdown in an otherwise stellar history -- nothing but gains -- would make an investment look terrible. Not sure this is the one-size-fits-all metric I've been looking for.
Actually, the metric does account for all of the years that are positive. It's basically a RMS of all of the losses from previous highs that divides by every year available (positive or negative). So a portfolio with only one down year out of 47 will definitely benefit from all the positive returns. You can read more about the specific formula directly in the original author's own words. Note that he also offers a simple spreadsheet download to show you how to calculate it for yourself.

nisiprius wrote:
Wed Nov 01, 2017 12:53 pm
"Ulcers?" Seriously?

I think it's been several decades since people believed that ulcers are caused by the anxiety of having to make difficult executive decisions. Nowadays they seem to think it's Helicobacter pylori.
:D To be fair, Peter Martin and Byron McCann coined the term "Ulcer Index" in 1987. Perhaps we can think of a better name today, but I wanted to make sure they got proper credit for their idea. Despite the medical inaccuracy, I personally think it gets the point across pretty well.
Last edited by Tyler9000 on Wed Nov 01, 2017 2:32 pm, edited 2 times in total.

User avatar
triceratop
Posts: 5838
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: The Ulcer Index

Post by triceratop » Wed Nov 01, 2017 1:24 pm

Tyler9000 wrote:
Wed Nov 01, 2017 1:20 pm
Thanks Kevin K. I'm glad you found the article useful!

livesoft wrote:
Wed Nov 01, 2017 12:26 pm
I was just having a discussion on this subject earlier today. I haven't really used the web site. Can one put in a ticker symbol and get an Ulcer Index? For example, VSMGX and VWENX?
That level of data integration requires an expensive subscription I can't afford. But as Triceratop illustrates, you can approximate the returns by modeling a portfolio with a similar construction.
Since you're here -- why do you not have any asset-class level international bond data?
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."

daveydoo
Posts: 1564
Joined: Sun May 15, 2016 1:53 am

Re: The Ulcer Index

Post by daveydoo » Wed Nov 01, 2017 1:29 pm

Tyler9000 wrote:
Wed Nov 01, 2017 1:20 pm

daveydoo wrote:
Wed Nov 01, 2017 12:51 pm
But those charts only include data on the depth and breadth of the drawdowns themselves, without factoring in all the drawdown-free time, right? One big drawdown in an otherwise stellar history -- nothing but gains -- would make an investment look terrible. Not sure this is the one-size-fits-all metric I've been looking for.
Actually, the metric does account for all of the years that are positive. It's basically a RMS of all of the losses from previous highs that divides by every year available (positive or negative). So a portfolio with only one down year out of 47 will definitely benefit from all the positive returns. You can read more about the specific formula directly in the original author's own words. Note that he also offers a simple spreadsheet download to show you how to calculate it for yourself.
I didn't get that from his explanation: "Technically, it is the square root of the mean of the squared percentage drawdowns in value." I don't see where the total drawdown-free interval plays a role in this. It seems to just compare drawdowns across investment, but I'm no expert.
"I mean, it's one banana, Michael...what could it cost? Ten dollars?"

User avatar
Tyler9000
Posts: 479
Joined: Fri Aug 21, 2015 11:57 am

Re: The Ulcer Index

Post by Tyler9000 » Wed Nov 01, 2017 1:34 pm

triceratop wrote:
Wed Nov 01, 2017 1:24 pm
Since you're here -- why do you not have any asset-class level international bond data?
The short story is that I model bond funds myself using data directly from various central banks. That's relatively straightforward for a single country, but to accurately model a multi-country international fund I need data (since 1970) for the proper annual market weightings of each country. If anybody can point me to such data, please let me know! :D

User avatar
Tyler9000
Posts: 479
Joined: Fri Aug 21, 2015 11:57 am

Re: The Ulcer Index

Post by Tyler9000 » Wed Nov 01, 2017 1:39 pm

daveydoo wrote:
Wed Nov 01, 2017 1:29 pm
I didn't get that from his explanation: "Technically, it is the square root of the mean of the squared percentage drawdowns in value." I don't see where the total drawdown-free interval plays a role in this. It seems to just compare drawdowns across investment, but I'm no expert.
It's admittedly a little dense when written verbally. But in the spreadsheet he provides (linked below) you can clearly see that the Ulcer Index divides by the total number of data points including every zero that indicates no drawdown at all.

http://www.tangotools.com/ui/UlcerIndex.xls

User avatar
siamond
Posts: 5131
Joined: Mon May 28, 2012 5:50 am

Re: The Ulcer Index

Post by siamond » Wed Nov 01, 2017 3:10 pm

As I indicated in the broader thread about PortfolioCharts, I applaud Tyler's move (and the excellent accompanying write-up).

MachineGhost, a very knowledgeable poster who shows up every now and then and then disappears, talked me in introducing this little known risk metric in the Simba spreadsheet (late last year, I believe). This being said, Simba works well for folks with Excel skills, but PortfolioCharts with its fantastic user-friendliness has a much broader appeal, and I am always glad to see Tyler make a push or another towards a good idea.

(to answer livesoft's question, somebody with some Excel skills could easily get the annual returns from Yahoo for a given metric, plug them in Simba, and get the Ulcer Index result -and more-, but this is admittedly NOT the neat and user-friendly interface that you probably had in mind! It is also relatively simple to compute it yourself from the returns, actually).

Now, as to the metric itself, I find it MUCH better than either std-deviation (SD) or max drawdown. The latter has the same issue as the SWR metric, it's a single data point, just one past singularity. SD spans the entire retirement period, but has many issues, starting by capturing ups (good!) and downs (bad!); Peter Martin's graph is just eye-popping, no need to say more. While the Ulcer Index does seem to neatly capture the full roller-coaster of negative emotions an investor can go through for an entire period of time (e.g. retirement or else).

This metric should be computed on a weekly or monthly basis to be fully significant, but after running some experiments, it seemed to me that the annual computation was a fairly reasonable proxy. Something that bothers me is that a very sudden peak at the very end of a bear market can cause a much deeper drawdown, and I don't think that emotions would be properly calibrated from the very top of the peak, it (should?) take at least a few months for a portfolio peak balance to stay deeply anchored in one's mind. But I'm probably splitting hairs here.

This is NOT a one-size-fits-all, risk is multi-faceted and one should never forget that. But the Ulcer Index became one of my preferred metrics to look at, at least for the emotional side of things.

2b2
Posts: 295
Joined: Sun Sep 28, 2008 7:21 am
Location: Location: Location:

Re: The Ulcer Index

Post by 2b2 » Wed Nov 01, 2017 3:19 pm

nisiprius wrote:
Wed Nov 01, 2017 12:53 pm
"Ulcers?" Seriously?

I think it's been several decades since people believed that ulcers are caused by the anxiety of having to make difficult executive decisions. Nowadays they seem to think it's Helicobacter pylori.
Nisiprius,

Can you provide the ticker symbol for that?

2b2

Post Reply