New User: Loving P123 - Question on Fundamental Data for ETFs

Hello to all,

A timid hello, just starting with P123. I’ve been lurking for a while, so am rather intimidated by the level here.

I’ve been on a free trial and am mighty impressed with P123, will be continuing with it. I sense the raw power under the hood.

It could be more user-friendly, though, both in UI (intuitiveness) and documentation (i.e., clarity of explanation and keeping the screenshots up to date). Paul, though, has been SUPER-helpful with my lengthy emails (big thank you, Paul!).

As an example, I must have spent a couple of hours just figuring out PctDev…

http://www.portfolio123.com/mvnforum/viewthread_thread,1341_lastpage,yes#37776

I just thought I’d dip a toe in a new thread here to say “hi” and to ask a question, a few questions actually, but all on one topic… ETFs

I’m more interested in asset allocation and ETFs than stock-picking. But there seems to be little conversation on that topic. Is that because few here are interested in that?

Or is it because most of the functionality in p123 is technical/statistical, with little in the way of fundamentals? It would be amazing, for example, to have historical data such as Price/Book for ETFs.

I know that I can download current ETF fundamental data from ETF.com (which used to be the same price as P123, now free, making P123 affordable! :wink: ) and also ETFdb.com. That is what I do, importing it into Filemaker, where I keep a master database of ETFs for quick access to a central repository of information, developing my own quality ranking system (not based on performance but on quality of the ETF itself).

But I can’t create backtests on that, nor on fundamental data (although at least I can screen on it). But it sort of defeats the purpose to create custom value screens in Filemaker and use that to create custom lists to replace what could have been fundamentally filtered here, and back-tested, etc.

Some fundamental historical data and screens would add so much.

Many think that there’s not a lot that can be done with ETFs and fundamentals, since Price to NAV (the value of an ETF) rarely develops much of a premium or discount due to the nature of the beast. And that’s true.

But, staying with the same example, Price of an ETF to the book value of its underlying assets WOULD be helpful. Even simple data like AUM and date of inception (I can’t find how to get that) would be helpful.

Given the relatively small number of ETFs compared to stocks, and that P123 has a massive collection of stock data, the “leap” to achieve this seems do-able? The software already exists, it’s just the data that is missing.

Is it that there is just no interest among users here?

I guess what I’m asking is whether anyone knows if there are plans to extend the “fundamentals” usefulness of ETFs in P123?

And, if not, does anyone know of any other good data sources with historical Total Returns, etc. If need be, I’ll import it all into Filemaker (similar-type scripting but incredibly usable) to cover the fundamental screens.

It would be nice to back-test these, but fundamental filtering probably has less downside that creating technical screens. And please don’t get me wrong…

I’m really enjoying the software, so I’m not complaining. I think/hope that I’ll make some breakthroughs here (famous last word?).

On the other hand, I can see myself hitting up against a ceiling at some point in the future. Still, if I can develop some intelligent screens that work, I’ll be happy.

P123 is much more flexible that ETFreplay.com, but I’ll keep using that, too, to set up quick tests, THEN get more creative here.

In short, I’ve loving the flexibility, working through the barriers and would love to know what is planned regarding fundamentals.

Thanks very much for any knowledge you can share. :slight_smile:

And it’s good to meet you.

All the best,
Ken

Ken, welcome to P123!

There is a significant problem with creating fundamental data for ETFs. Not necessary for current fundamentals, like you say we have the stock data so P123 can download the list of stocks that make up an ETF and create averages of the fundamental data. But, and it is a big BUT, it is very difficult if not impossible to get an accurate list of stocks that were held by an ETF in the past.

Without an accurate list for each date in the past, and we just used the current list in the past there would be a significant look ahead bias. That would make the value of running back tests iffy at best. P123 may be able to get a historically accurate list for a few of the ETFs like DIA, SPY, QQQ, and IWM. But I fear we couldn’t get data on a large enough set of ETFs to make a good trading system.

Don’t get me wrong. I would LOVE to be able to run simulations of fundamental ETF data. And I hope that it can be accomplished on a large enough set of ETFs to make the effort worthwhile. We need P123 to chime in on what they see as the possibilities.

Denny :sunglasses:

Thanks very much, Denny.

I did not consider that the lists of ETFs that make up a fund, varying since inception, would not be available. Considering what IS available in those massive database, it seems almost trivial.

Too bad.

Thanks again.

All the best,
Ken

P.S. I’m a fan of yours. Several of your answers have helped me out. Your posts tend to come up fairly often when searching. So thanks, too, for those! :slight_smile:

Well…

First the current constituent list is not that easy to come by (certainly not from S&P/Compustat. They only recently added an ETF taxonomy, which is why we did it by hand).

Second, historically correct constituents data we don’t believe exists.

Third, there will be constituents in ETFs from other countries so we would need World data.

You put all those together and it’s a massive^2 project. The only way to recoup the investment of such a project would be to resell the data.

PS. Reuters had some ETF fundamentals (current, no history). They were incredibly spotty, mostly N/A’s. Only the main etfs had them, making this data useless in our view. And… when we asked for details, nobody really knew how or what was generating them.

Hi Ken, Denny, and Marco,

I can totally appreciate Ken’s question, as I created an ETF-only turnkey website many years ago in anticipation of P123 cracking this $3 billion/year sub-industry, but I have yet to launch the site. The reasons that Marco gave are very sound, and are very accurate from my personal research of the available data.

ETFs are very appealing to both pros and amateurs, as they are a convenient way to participate in - or hedge against - broad market moves (whether sectors, domestic, or international). But there is no good way to backtest the existing data, other than technical analysis.

I have a 3X ETF portfolio based on a single trade in-and-out of TNA that is producing a return of 147% annualized, and a similar 2X ETF based on SSO that is producing a return of 72% annualized. That’s not too shabby a performance, given that the average investor earns less than 6% annually on their money. However, those trades are based on going to cash (hedging isn’t available for ETFs yet) when the market goes against us.

ETFs were a hot topic following the 2008 crash, as many saw them as a way to have a single trade just a few times per year (or decade), and they would profit regardless of the direction of the markets. I’m glad to see that ETFs are still of interest to users here, and hopefully, somebody someday will be collecting the constituent fundamental data and we’ll be able to pick individual funds based on value, rather than price performance. In the meantime, using technical analysis (I recommend oscillators, not moving averages) may be our only current option.

However, Marco has proven to be very resourceful over the last 10 years that I’ve been a member, so who knows what he will find/create if interest is shown. So chime-in with some comments to Ken’s post, ALL! Exchange Traded Funds are the wave of the future!

I would echo the general theme of Ken’s post. It seems as if P123 has enormous computing power and is continuously introducing new factors, data, and now R2G ideas (autotrading, etc.), ETF features in general seem to seriously lag features added to the stock databases. In addition, position sizing and asset allocation functionality in its most basic forms has been left far behind the other features.

Bottom line, if fundamental ETF data is too big of an undertaking, additional asset allocation tools would be of a major benefit to the ETF/CEF side. Volatility weighting (risk parity), minimum variance weighting, correlations matrices, are all possible using data already available to P123.

If I had a guaranteed 147% trading TNA I would chuck in 500k and be worth $45 million in 5 years or $4.2 billion in 10 years and I wouldn’t bother peddling it to anyone else.

That’s the plan, Gary. Like many on here (see the weekly performance report), I have stock portfolios producing double that. But like Buffett did when he launched his first partnership, he used other people’s money to get started. If you’re truly interested in doing what you said, please contact me and I’ll consider your offer.

Looks like that was also the plan in 2007/2008 under a different portfolio name Review of IntelligentValue's Retracement-Value Portfolio - CXO Advisory

"With the magic of compounding, when I add an extra $5000-$8,000/month of the profits to my Retracement-Value-based portfolio, my wealth grows at an enormous rate. An investment of $10k over 10 years at 200% becomes $590 million. With a monthly addition of $5,000, that same $10k becomes $5.9 billion. "

Scott (oyenscot),

Yes, that’s correct. It was the plan then and it’s the plan now and for the foreseeable future to make money and accumulate assets. What’s your plan?

I think this post is getting off the subject of why the ETF section should be developed more and on the subject of why I publish a newsletter. My record has been plain for anyone to see since I launched my site in 2004. I have several portfolios, and the results are documented daily/weekly. My documented performance is not that impressive (124%-220% for public portfolios) because I have to cater to the general public/individual investors - who are subject to emotional stresses that are counterproductive to their financial well-being. Therefore, I have to avoid drawdowns in the public portfolios - at the expense of long-term returns. There are many, many other system developers who are producing returns much greater than the ones I’m publishing.

My contribution was based on the idea that I would chime in with the voices wanting to see the ETF product developed more fully (hedging in particular). I also wanted to urge other members to speak out to let Marco know it’s desired. Instead I get flamed for wanting to make money and be wealthy. Very impressive behavior.

By the way - CXO Advisory is a site by academics dedicated to telling people why they can’t make money in the stock market. I’m surprised a fellow investment publisher like you, Scott, would be promoting their BS.

Hi to all,

I didn’t mean to spark such an interesting discussion, but am delighted that so many luminaries here commented. I thought that ETFs were really the neglected baby of P123, perhaps a lost cause, but some obviously do care.

Personally, my sole interest is in ETFs and optimal asset allocation using sound, customized analysis based on findings in academia. Much has been overlooked still and can be derived and tested (well, once I master this tool - I’ll get there, believe it or not).

It’s a shame that so little historical fundamental information is available. What is here (on the fundamental side) is of little use. There are, though, many useful options on the technical side for those focused on ETFs.

A few comments to the posts here…

  1. It boggles the mind that this youngest, most electronic field of investing, has no digital record of the day-to-day composition of each ETF. Clearly, it’s not possible. But in this over-regulated world, where everything has to databased somewhere, I would have never thunk it.

Given that, though, I have a question regarding what course of action you would be willing to take

Since a combination of value and momentum screens are known to be more powerful than either alone, AND since CURRENT fundamental data for ETFs can be downloaded…

Would you take an existing momentum screen that is your “current best” and use it in conjunction with, say a P/B and P/E ratio screen using today’s values (or basic bond ETF parameters, etc), without any backtesting (assuming, of course, that you do actually find some fundamentally-promising ETFs among the Asset Classes that are in your technical ETF screen)?

Absolutely not, never? Or maybe, depending? So a small test? It’s just that I have a hard time leaving fundamental values on the table

2) Scott, I agree with you (re what ETF users need). Any help at all, such as the ones you list, would be wonderful additions for us ETFers.

By the way, I’m a big fan. It’s one of your articles that has me here. :slight_smile:

  1. Marco, I know that you have to balance the needs of 2 groups (with one that is much larger) vs allocatable resources. And if most folks are not interested in ETFs, it would not make sense for you to develop this side a great deal further. It’s obviously much quieter here.

But if you can implement some of Scott’s suggestions (and I’m sure you have a wish list of your own that would add value to “value”), even adding the simplest ones would be a great start.

  1. Christopher, I don’t think anyone here is flaming you. Everyone here (excluding myself from this group – honestly, I’ve never felt inferior before!) is obviously extremely strong in statistics and knows this field inside out.

So you suffer, a priori, from the “looks like a duck and quacks like a duck” syndrome. Meaning - the odds are high that results are overstated and the sales copy is aggressive.

CXO did a dispassionate analysis, in their typical fashion. Over and over, they simply call it as they see it, backing up what they say with logic and data.

Best of all, though, is that they’re open to responses. But over and over, people get defensive instead of rolling up their sleeves and giving them what would impress them most – complete and well-documented data. That’s what drives them.

They are not the enemy.

The worst thing to do is reply with threats of a lawsuit. The best thing would have been to sit down and NOT beat them, but convince them.

For that, your system has to stand up to full scrutiny.

That said, you do deserve a fair shake. The quiet presumption of guilt lingers here, I don’t disagree. Even though it’s based on statistics, it is still not fair, I agree. (There may be history that I’m not aware of, but, if so, let’s leave that out.)

I’d love to see someone here PM you and offer to sign an NDA if you allow that person full access to dig into the entire system, then publish a report here and to CXO. If I was you, that is what I would offer.

Heck, if no one else does, I would. I do have a worry, though. The world should be beating a path to your door. But traffic stats are non-existent…

General Traffic (no data means traffic is lower than the 25,000,000th site, which is non-existent, basically)
http://www.alexa.com/siteinfo/intelligentvalue.com

Search Traffic
http://www.semrush.com/info/intelligentvalue.com

If you are not actively marketing this anymore and if you are just offering it privately here, I would study this opportunity if no one else wants to.

Please PM me if you get no offers, but do know that while I may be slow at learning P123, I’m otherwise meticulous and get to reality sooner or later, meaning that I’d go into it deeply. If I can understand this completely and accept the concept, I’d put at least $100K of speculation money into it, but not serious investing money unless it subsequently proves itself.

But, back to your efforts. The perception of flames is defensiveness. You have to realize that it sounds too good to be true. The sales copy does not help allay that fear. But while folks here (including me) may start with pre-conceptions based on likelihood, we never know for sure until we lift the curtain. :slight_smile:

Hopefully that was reasonably balanced. There is a very clear way to turn this into lemonade. :slight_smile:

And now as you mentioned, THAT discussion is pulling this off topic. So, getting back ON…

  1. I have found myself wanting to be able to call a simple date of inception to rule out ETFs prior to a certain date to be sure that no ETFs (including lookback period) run into the dates of back-testing (Marco, I have emailed Paul, in great detail, with what I believe is a bug in that regard).

Even having AUM would be great.

Also, some low volume ETFs can be very liquid, unlike stocks. so a volume rule is not the best for ETFs. I don’t know if bid-ask spreads are difficult to get or not (although even those are not 100%, but it would get us closer).

Scott’s bigger ideas are great. If one or more are do-able, I too would love correlations. Oh, and here is a big (but easy, I think) one…

We need to be able to indicate our own benchmark. The choices are too limited. It could be as simple as entering any ETF symbol (or formula) to be the benchmark (there is an ETF proxy that tracks just about any index out there well enough, and sometimes we just want to compare against a solid ETF).

To other ETF investors, please suggest what would help you, items that are not cost-prohibitive, understanding that it does not make good business sense for a business owner to allocate resources to ETFs disproportionately unless he wants to make a bet on the future.

But, with all due respect, that’s his business. All we can do is ask for the most useful data or features that take the least amount of time and money.

Thanks very much for giving this attention. :slight_smile:

All the best,
Ken

As the main initial advocate of ETFs on p123, and as the guy who actually does the manual taxonomy classifications to which Marco referred (so if you find an error, I’m the one who gets to wear the dunce cap), I figured I should chime in on this topic.

P123 is an equity platform, not a fund platform.

That statement is far more substantive than many might realize. We are powered by equity databases (Reuters until about a year ago and now Compustat). We do not now and never have licensed fund data (whether open-end funds, ETFs or closed-end funds). Fortunately, however, the boundaries between equity data and fund data are not always iron-clad. For example, the vendors from which we license price data include ETFs and CEFs in the content they supply us. That opened the door to our adding ETFs at first and later CEFs – assuming we were willing to work with what we could get (price and volume data). As time passed and as we changed equity and pricing data providers, we discovered, much to our joy, that we could add a couple of other items; hence or ability to offer yields and for CEFs, NAVs (from which we can derive premiums and discounts). Ultimately, though, and notwithstanding the aforementioned episodes of good fortune, the fact remains that p123 is an equity service limited to content supplied by our equity data providers.

Hence fund-oriented content such as inception dates, AUM, expense ratios, portfolio manager identify and tenure, payment drill-down (dividend or return on capital) is off the table unless we choose to make a major change in the nature of p123; i.e. unless we decide that in addition to being an equity platform, we will also become a fund platform (in which case we’d have to negotiate licenses with one or more fund data providers such as Morningstar or Lipper). Whether we go there will, like anything else, be a business decision; i.e. based on our assessment of potential demand, incremental subscription fees fund investors would be willing to pay, cost of licensing fund data, costs associated with the considerable development man hours we’d need to devote to incorporating fund data into our platform (including point-in-time), etc.

Especially important would be the benchmark against which such business analysis would have to be measured. We are not analyzing the choice between offering nothing on funds versus being a fund platform. Instead, the choice is between being a bona fide fund platform versus being what we are now viz. funds; a “lite” fund platform that is actually able to offer quite a bit on funds, but not all such an investor might want.

From the point of view of p123 members, the question is how much extra return you could generate from such items as AUM and expense ratio versus how much extra you’d be willing to pay. At this point in time, we’re not seeing any indication that it would be worthwhile to pursue the measures it would take to make p123 a bona fide fund platform as opposed the funds-lite we offer now.

Bear in mind, too, that it’s not a choice between devoting many hours to funds versus sitting around watching daytime TV. We have plenty on our plate in our core equity competency. So either way, we’re quite busy. The question is what we choose to work on.

Now, as to what we do offer re: ETFs.

The idea that motivated us to get into this was the desire to offer p123 members more choices regarding investment exposure. Before we took the plunge, all we could offer long exposure to U.S. equities. ETFs instantly enabled us to offer exposure to fixed income, leverage, shorting, global equities, commodities, currencies and hedging strategies. Since then, we’ve added leverage and shorting to the regular screener/sim, but for many, leveraged and short ETFs are still a preferred way to gain such exposure since they involve no special brokerage-account baggage. We’ll eventually offer global equity coverage but that’s a long process and it, too, involves brokerage account issues.

As to the idea of selecting ETFs based on fundamental characteristics of the stock in the portfolios, first, let’s understand that portfolio composition, if available at all, is fund content and is, hence, off the table as noted, unless we decide to become a bona fide fund platform. But even if we could snap our fingers and do this, let’s slow down a minute and think about what incremental benefit we’d actually be getting. First, right off the bat, we’d be eliminating 63% of the ETF database (as of today, there are 569 US equity ETFs out of a total of 1,547 ETFs) and pretty much all the new exposures that motivated us to get involved at all with ETFs.

Assuming one would be willing to wipe out 63% of the ETF database and most of the new exposures that are the main benefit of our having added ETFs, let’s go further and see what unique benefit we’d actually get by adding the ability to fundamentally analyze ETF portfolios. Say you want favorable valuations, strong balance sheets, good returns on equity, and a dash of recent share price momentum. Do you really think you’d be better off applying relevant model logic to US equity ETFs than simply building your own portfolios (personalized ETFs, so to speak) that give you exactly what you want? Might you be concerned that transaction execution/burdens could overwhelm you? Then trade through a FolioInvesting.com account; you’d be limited to their 11 AM or 2PM trading windows, but that still gives you far more control than you’d get by owning an ETF. (Do you really know when and at what time of day they rebalance their portfolios?) If you care that much about the fundamental characteristics of your domestic equity exposure, then why mess around with ETFs at all and pick from a very limited palate those ETF(s) that seem to come closest to what you want (but still may miss by a country mile). If you really care about fundamental characteristics of your domestic equity exposure, then you’re right in p123’s wheelhouse, where we’re a significant power. Use our core platform and get exactly what you want.

ETFs are for those who don’t care so deeply about such bottom-up considerations. ETFs are a top-down product line. That’s the framework you need in order to be investing in ETFs rather than stocks. And it’s a perfectly legitimate framework (as is evident from the overall success of ETFs). But there’s nothing fundamental data on holdings (a bottom-up thing that can be much more effectively handled on p123’s regular equity platform) can contribute to such a framework.

The bad news, so to speak (not just for p123 users but EVERYBODY who invests top down) is that the field remains very primitive. We have lots of opinions about asset allocation, but little in the way of credible methodologies. Markowitz optimization is theoretical masterpiece but a real-world mess with “portfolios” being typically “dominated” by a single security with extreme historical risk-reward data trends, which tend to be used (notwithstanding to absence of serial correlation) because it’s easy and because nobody really has a better way to come up with inputs for expected return, volatility and correlations. (The Black Litterman approach is a bit better but still leaves much to be desired.) Other CAPM or APT approaches likewise suffer from inability to come up with credible inputs. A lot of what is seen out there follows no methodology at all. Target date funds, for example, are based solely on folklore and stereotype. (Do we REALLY think those nearest to retirement and who presumably should be most risk averse ought to load up on fixed income right now given that we’re probably at or near the top of what could turn out to be a major bond bubble?) Folklore likewise guides much else we see in the area of asset allocation (there’s lots out there on “what” people should do but little credible discussion of “why” other than reliance on historical correlations with little regard to the fact that correlations lately have been diminishing not just as a matter of statistical circumstance but fundamental developments).

I like to think that over time, we’ll get better at top-down investing and it would not surprise me at all if we were to ultimately learn that those who make notable progress in this area turn out to be p123 subscribers. We’re already a domestic equity powerhouse, and we just dipped our toe into the area of economic data and will be doing a lot more over time. I strongly suspect that use of sophisticated economic data to model against the full range of ETF offerings (not just the 36% from which a fund data vendor might – not “will” but “might” be able supply to supply current and point-in-time portfolio—holdings fundamentals) is the route that will eventually lead to the solving of the sort of top-down analytic challenges that will eventually drive ETF strategies. And as to the timetable for beefing up our economic capabilities, that goes back to what I said earlier. We’ve got a full plate, and if you really are interested in serious (as opposed to seat of the pants stereotype-based) asset allocation, the last thing you should want is for us to divert resources to things like ETF expense ratios, AUMs, etc. and away from projects on our to-do list that can be real difference-makers to you.

In the meantime, the technical analysis and taxonomy choices we presently offer should not be taken lightly. Not everything has to be automated to the nth degree. You can do a heck of a lot today and as time passes, you’ll ba able to do a heck of a lot more.

Hi Marc,

Thanks very much for taking the time to elaborate your thoughts on ETFs. They confirm what some of us needed to know, that ETFs are an incidental “because we could” add-on, not a core offering that will receive much development. My takeaways from your post…

Bottom line… There are not enough “EFT-only” investors to justify it. The ETF data delivers added usefulness to your core market, stock investors, and incidentally also provides some usefulness for those of us who are more interested in ETFs.

Understood. So there is no point in asking for anything that requires new data expense or considerable programming. I make the same decisions, too, in my business, and would make the same one that you have. And that’s OK…

Even “as is,” P123 does provides genuine value to those who are only interested in ETFs and who want to explore certain theories more deeply. ETFreplay seems to be the only other game in town, and while its interface is intuitive and its range of pre-set functionality is excellent, it does not enable one to customize beyond pre-set drop-downs.

Both have their particular usefulness. I wanted to push beyond what I was doing with ETFreplay, hence P123. I can also rough-quick-test some methodologies laid out in an academic paper faster in ETFreplay than in PI123, then decide whether to go deeper here. So they are both useful, each in their own ways.

For any who feel like me, we may be disappointed at not having more extended functionality. But a well-run company that does not chase the wrong markets means the continued availability of this excellent tool. :slight_smile:

I don’t think, though, that your argument needed to go further beyond your business case. You lost me once you went there, first because we are now into a hypothetical that has been ruled out, and second because it reads as argumentation that is uncharacteristic of your writing, more to make a point than to present a fair and balanced argument regarding the potential value of the requests (assuming you could, as you said, “snap your fingers and make it happen”).

I agree with you that many ETFs, especially the newer ones, are merely marketing vehicles that add little (ex., “target date” ETFs for targeted demographics). And I agree that ETFs are top-down. But neither of those points obviate the benefits of having both high-level fundamental AND technical data upon which to screen.

Nor does the suggestion to use P123 to build and maintain our own rules-based ETF-equivalents (which is hardly as trivial a matter as you present it to be). As you said, we work top-down. Recreating 100s of ETF strategies from the bottom-up, just to test them, would be not only laborious, but artificial.

Some of your points selectively try to make cases that are actually more nuanced than presented. Yes, some asset correlations are converging. However, several macro-driven classes (ex, stocks, long-term treasuries, and gold) remain as non- and negatively-correlated as ever. The underlying drivers remain the same - prosperity vs contraction, high-interest vs. low, liquidity, crisis conditions, etc.

And I disagree with other points, such as the “even if we did this, there’s little benefit” argument. Your discussion of this shows that you look at this through a very different lens, missing many possible creative and barely explored (even in academia) possibilities, concepts that are potentially more powerful than what has been demonstrated anywhere, to date. The ability to back-test these new approaches, even if limited to the recent history of ETFs and hence not at the level of academia, would be most useful - I would invest in the most promising discoveries (needing only a sufficient cushion for a level of confidence).

I’ve shortened my reply and therefore oversimplified (hopefully not at the expense of over-bias), because there is no point in a detailed discussion about a rejected hypothetical. Your business bottom line remains the same…

Even if I am correct, there is nothing here that would not suddenly expand your market, leaving no reason to change your “no business case for this” answer, which I fully understand and respect.

I merely felt that some of your “post business case” arguments needed at least a bit of rebuttal.

The business case conclusion is what is it is, and it suffices. Every company has too many good ideas than they could/should ever implement, even Google and Apple. All we can do is allocate resources that best meet our core customers’ needs.

That, I believe, was your real bottom line.

And getting THAT right is hard enough. Rejecting goods idea does not mean that you or I need to feel that they were bad or useless ideas. There is simply more to gain by choosing to go with what your core market needs if the potential of a new market is not large. And on that note…

This thread is more powerful proof that your business decision is sound. Few people, only those interested in ETFs, spoke up here. Case closed. :slight_smile:

Recognizing reality, please allow me to reduce the request to one suggestion which has no data cost and minimal programming time…

Benchmark of ETF Backtesting

The current benchmarks that are available cover pretty much every style, size and sector of equity, as well as all the basic indices. In a nutshell, stocks are well covered.

But ETF backtests cannot benchmark, say, a momentum strategy for gold against the price of gold (ex., the London PM Fix) – there is no such option in the Benchmark dropdown. Assuming you do not have access to that data, and since getting it would entail an additional cost, and since there are several other classes and sub-classes like this, here is what I suggest…

Please enable the user to benchmark against any ETF, IAU or GLD in the example above, by entering the symbol into a text entry box.

That would more than suffice. There are no new data costs, and only minor dev (since you will be using routines that are already programmed).

ETFs cover much more than stocks – other asset classes (ex. commodities, fixed income, etc.), specific countries and regions, methods, etc. Please let the user determine the single ETF proxy that best benchmarks the class, sub-class or even mix of classes that is being studied.

Benchmarking is a critical part of backtesting. Studies lose a great deal of their information and purpose without that ability. But we can’t do it for most types of ETF screens and their backtesting.

Thanks very much, Marc, for taking the time to express the views of P123 so clearly, and for considering the only request of significance that remains, one that costs very little, but would add much for “ETFers.”

With appreciation,
Ken

P.S. This comment was very interesting…

Being able to choose a proper benchmark would help. :slight_smile: And yes, I do believe that we’ll come up with some original work. Scott and others are way ahead of me in that regard. There are some really creative and smart folks here.

But judging from the academic literature, 20-40 years of backtesting seems to be the minimal requirement in order to have a high degree of confidence in the results, which means publishing the statistical analysis to support one’s conclusions. But that raises a problem…

There are not enough ETFs with the longevity to cover full business cycles or to do tests that deliver higher levels of statistical significance.

So no matter what we develop with P123 for ETFs, there’ll be potential blind spots. That said, should we find an approach with a large enough “alpha cushion,” it’s worth putting into action at the practical level.

Card counting was being successfully used well before E.O. Thorpe, after all. It’s nice, though, to know you have the math backing you up. :slight_smile:

So…

The development of recognized notable progress would require the presence of longer-term historical total return data, volume, and fundamental data on a range of indices that extend beyond U.S. stocks to such other asset classes as developed nation equity, long-term treasuries, credit risk bonds, real estate, t-bills, London PM gold fix, etc.

That would enable some truly ground-breaking work. I know that is even less of an option for P123. But it’s a direction that interests me…

Academics develop these systems using R, serious modeling software (Excel is also used). A programmer friend of my daughter, who used to program for a major financial firm, is looking to develop such a system for me in R, but it’s early days and I’m sure there are many pitfalls along this road. If anyone reading this has any interest or expertise in this, please PM me?

NOTE: Edits made to balance open and close tags of italics and bolds, fix minor typos. No change in content.

Ken,

There have been a number of feature requests over the years to add additional ETFs to the benchmarks.

One thing to keep in mind is for an ETF to be most useful as a benchmark in back testing it needs to have been started in 1999. That significantly limits the number of ETFs that would work best. P123 has extended back in time some ETFs that had a start date after 1999 by using the ETF’s index as a proxy for calculating what the ETF performance would have been back then. However, that also requires that the index existed in 1999.

Denny :sunglasses:

“This thread is more powerful proof that your business decision is sound. Few people, only those interested in ETFs, spoke up here. Case closed. :-)”

Ken et al - I don’t think it is right to conclude there is no interest in ETFs based on the response in this thread. The desire for top down design starting with ETFs has been voiced before. The topic of this thread was “fundamental data for ETFs”. I don’t believe this is practical and I don’t feel it is of that much benefit so I didn’t respond. This doesn’t mean however that we don’t need a top down plan involving ETFs.

I’m not sure about others but I feel like a sitting duck holding stocks when a bear market appears or is apparently just around the corner. The fear that this brings on makes for poor decisions at the account level i.e. closing a port when it is exactly the wrong time to do so, etc. The current hedging module has too short a time frame and too few instruments / logic to make it practical. It is almost dangerous the way it is because we can only model over a short time frame.

So some of this is being taken care of (I believe). I understand that a better hedging module is in works as well as a long time period for economic indicators. But I do believe there is a need to be able to integrate ETFs and stock portfolios at the top level. What I mean by this is to be able to generate an asset allocation model based firstly on indices, followed by ETF substitutes for the indices, then replacing individual ETFs with stock portfolios where possible. There is no need for hundreds of ETFs, probably a dozen or so. We just need to be able to cover basic asset classes with ETFs that are fairly representative of long term indices (like the spot price of gold for example).

Is there a business case for this? Ultimately, it is P123s decision, but the way I see it is that right now only geeks (mathematic/finance) come to this site, perhaps a few R2G followers. I believe a top down methodology implemented in a simple fashion could draw a wider crowd.

Steve

Hi Denny,

Yes, I agree with you, that would be ideal (1999). But “less than ideal” is still better than nothing, and I believe that the ETFs that have been extrapolated back have all been equity ETFs, so there remains no way to benchmark for screens outside of the realm of equities.

I’m happy to settle for less than ideal benchmarking (vs no benchmarking) in non-equity screens. And, per Marc’s clear business decision, I’m trying to keep my request in the form of “no new data needed” and “minimal programming.”

For example, if I could use an ETF to benchmark, say, an experiment with real estate, using VNQ as the benchmark, that’s better than nothing.

Or, picking up on Steve’s point about emotions causing bad decisions, suppose I wanted to set up a screen for single-country ETFs with the lowest P/B ratios. Every country would likely scare us and keep us out of them.

But if I could run a “10 countries with the lowest P/B” screen from 5 years ago, but that also had positive momentum, it may show such remarkable results that I don’t feel compelled to go all the way back to 1999.

But what would I benchmark it against? Well, if half of the countries were emerging nations and half were frontier, I could set up a formula of 50% EMB and 50% FRN, allowing me a benchmark back to June, 2008.

I could extend my selection back to the same date. The differences in CAGR, Volatility, MaxDD and Sortino, if negligible would leave me with little confidence – my losers would seem to have merely gone along for the ride.

But if my “formula benchmark” had dropped while my losers had jumped, I may only have a 6 year test, but the chances that this difference is random is low (and can probably be measured?).

This latter example is the dream scenario is just one super-simple example of what would be possible with top-down fundamental screens. But I’m not asking for that data anymore, just the ability to benchmark ETF screens by specifying an ETF (or ideally, now that I think of it, a formula of ETFs).

I mean, the SP1500 Gold index is simply not going to give me ANY information about a screen involving gold. Surely being able to benchmark that against a Gold ETF is better than what we have now for ETF screens?

Steve, yes, my last post changed direction somewhat after Marc explained the business rationale for minimal development of the ETF side. But I think you’d be surprised at what fundamental screens could find, especially in combination with proven technical ones.

But let’s put that aside. It’s not to be.

And, if I understood what you were saying, your reason for not responding to this thread stems from being primarily a stock investor, interested in ETFs only as a potential hedge vehicles for your stock portfolio.

Those, though, who are interested in “ETFs as ETFs,” such as Scott, would be more likely to guess at where I was going. Or it might stimulate ideas of their own that build on the request, but always for ETFS in their own right, not how they can be used with the “stock side” of P123.

Your reason for not posting is actually consistent with my interpretation of the disinterest. I hope that makes sense, since I’m barely holding onto this conversation by my the tips of my fingernails. :slight_smile:

I agree with you that a nice selection of index data would be really helpful for both stock pickers and ETF investors. Meanwhile, there’s a ton of good academic papers, most of which use indices, about how you could hedge your carefully built stock portfolio if you’re worried. With so much systemic instability built into the system, I’d be worried, too.

For example, if you are worried about stocks, check out Gary Antonacci’s 2014 paper on Absolute Momentum…

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2244633

The paper was not written for that reason, but could be used to accomplish what (I think) that you are getting at.

Most of the gain from momentum strategies like this comes from holding their own, or even gaining, during DOWN (stock) markets, while basically tracking UP markets. It’s what I’ve found and what Gary has confirmed to me in email.

Hope this helps a little - I feel like an elf among giants.

But, unless I’ve missed something so far - for ETF investors, the ability to indicate one’s choice of any ETF, or better still a simple formula, that would act as a “control” (perhaps a better word than “benchmark”) against which we could measure the outcomes of screens - would be a big step ahead compared to what we can use now to benchmark.

Not perfect, just better. :slight_smile:

All the best,
Ken

“your reason for not responding to this thread stems from being primarily a stock investor, interested in ETFs only as a potential hedge vehicles for your stock portfolio.”

Not as a hedge. I can build all the stock ports I want and throw in some hedges. But I’m not seeing the forest for the trees. What I want is a high level “plan”. I want to take a set of long term indices that represent various asset classes/categories or sub-categories, etc. and combine them with formulas to come out with a long-term strategy for maximizing profit and consistency of profit while minimizing risk. But indices can’t be traded. So then ETFs come into play. Do I need fundamental data? No - I just need ETFs that closely match the indices. Then I can start thinking about replacing equity ETFs with stock portfolios.

** EDIT ** In summary, I’m not interested in ETFs as hedges for stock portfolios. I’m interested in using stocks to spice up the returns of a well diversified ETF-based portfolio.

Steve

Yes, yes, yes . . . wishing for the ideal (which never really happens) is a perfect prescription for paralysis. And I’ll even raise you one: I don’t go back to 1999 even though I can. As others may recall from prior posts, that’s just plain bad sampling. (A little but good sample trumps a big sample loaded with garbage.)

I’m not going to elaborate now because forums, where posts quickly scroll into oblivion, aren’t ideal for such things. But I eagerly await completion of the Channels project Marco recently mentioned on the boards. Paul and I have been hammering out requirements and Marco has a plan to get this developed pretty quickly. That’s where we can more effectively disseminate and discuss bigger ideas. Stay tuned; it’s going to be great stuff.

Meanwhile, I want to backtrack to the discussion of asset management etc. I’d love to see us get into that in a serious way. Frankly, though, the stuff I’ve seen is great for textbooks and classrooms, but pretty poor for real-world use. If anyone knows of approaches that really work under real-world conditions, please, please, please let us know . . . in the forum or privately if you prefer. That would amount to a bold step forward in working with ETFs, or combinations of ETFs and stocks. I do have some thoughts that evolve from use of economic data (as has been mentioned in the past, think of what we now as a starter set – more development here will occur). But we’re all ears for any other ideas.

Ken,

There are many more ETFs that go back to 1999 than equities. Here is the list:

Ticker Name
DIA SPDR Dow Jones Industrial Average ETF
EWA iShares MSCI Australia ETF
EWC iShares MSCI Canada ETF
EWD iShares MSCI Sweden ETF
EWG iShares MSCI Germany ETF
EWH iShares MSCI Hong Kong ETF
EWI iShares MSCI Italy Capped ETF
EWJ iShares MSCI Japan ETF
EWK iShares MSCI Belgium Capped ETF
EWL iShares MSCI Switzerland Capped ETF
EWM iShares MSCI Malaysia ETF
EWN iShares MSCI Netherlands ETF
EWO iShares MSCI Austria Capped ETF
EWP iShares MSCI Spain Capped ETF
EWQ iShares MSCI France ETF
EWS iShares MSCI Singapore ETF
EWU iShares MSCI United Kingdom ETF
EWW iShares MSCI Mexico Capped ETF
GLD SPDR Gold Trust
HHHYL Internet HOLDRS Trust
IEF iShares 7-10 Year Treasury Bond ETF
IWB iShares Russell 1000 ETF
IWM iShares Russell 2000 ETF
IWV iShares Russell 3000 ETF
MDY SPDR MidCap 400 ETF Trust
QQQ PowerShares QQQ Trust Series 1
RWM ProShares Short Russell 2000
SDS ProShares UltraShort S&P500
SH ProShares Short S&P500
SHY iShares 1-3 Year Treasury Bond ETF
SPXU ProShares UltraPro Short S&P500
SPY SPDR S&P 500 ETF Trust
SSO ProShares Ultra S&P500
TLT iShares 20 Plus Year Treasury Bond
TNA Direxion Daily Small Cap Bull 3x Short
TWM ProShares UltraShort Russell2000
TZA Direxion Daily Small Cap Bear 3X Short
UPRO ProShares UltraPro S&P500
UWM ProShares Ultra Russell2000
XLB Materials Select Sector SPDR Fund
XLE Energy Select Sector SPDR Fund (The
XLF Financial Select Sector SPDR Fund
XLI Industrial Select Sector SPDR Fund
XLK Technology Select Sector SPDR Fund
XLP Consumer Staples Select Sector SPDR
XLU Utilities Select Sector SPDR Fund
XLV Health Care Select Sector SPDR Fund
XLY Consumer Discretionary Select Sector

Denny :sunglasses:

Thanks very much for the feedback. Some replies…

Marc, could you speak to the one ETF screen feature request that remains, please? It has no data cost, minimal programming cost and, even given those constraints, would deliver a great deal of benefit to ETF backtesting, which is to say…

Could you enable ETFers (who are elaborating ETF strategies) to choose any ETF (or formula) as a benchmark when back-testing, simply by entering its symbol into a text entry box? Right now, stock indices from the drop-down menu are the only option. So the usefulness of backtesting any Asset Class other than equities is greatly reduced.

A superior option (to entering any ETF) would be the ability to compare the test screen vs. another screen, choosing from a list of existing screens. That way, the user could develop his own “benchmark screens,” which could be either…

  1. conventional “single-ETF” Universe (i.e., the conventional benchmark where one would choose one passive ETF that serves as a good proxy for an index)

-or-

  1. more involved screens developed specifically for benchmarking (ex., comparing an involved equity momentum screen vs, for example, a simple absolute momentum equity ETF screen, such as comparing the Relative Strength of VTI to that of SHY, moving into SHY when it is greater than VTI, otherwise in VTI).

The latter would be amazing since we could could see how much, and what type of, additional benefit is delivered by the more elaborate test screen. And yet, it requires no new routines, merely calling the “benchmark screen” and reporting/charting it, along with the test screen.

Custom-benchmarking would be amazing for the ETF side of P123 for ETF-only users. Not even ETFreplay does this.

With respect, could you speak to this request, please? Is it sufficiently low-cost and low-time-to-dev to consider?

Steve, your “edit summary” clarified perfectly…

Earlier, you had talked about feeling “like a sitting duck holding stocks” and about the inadequacy of the hedging module and then suggesting…

I didn’t realize how far you wanted to go, didn’t realize that you were looking for full diversification. You note that you’d have “no need for fundamental data,” which is true…

That is consistent with the point that I was making – this thread indicates that few in P123 have any interest in finding optimal ACTIVE ETF strategies, which shows in low level of interest in this topic (although it’s raised many other interesting directions :slight_smile: ).

The topic of asset allocation to protect your P123-built stock portfolio is a good one. I’ll post in a new thread about that, to avoid taking this thread too far off-topic. Hopefully, Scott will notice and help out.

Marc, re your comment on how many years of back-testing is enough, you said…

A little but good sample trumps a big sample loaded with garbage.
[/quote]

Can you point me to the URLs of those prior posts, please? I agree that bad sampling is bad (what else could it be?), but I don’t see how that establishes that short-term back-testing is good.

I would really like to see the reasoning. If it’s sound, I’ll be much closer to putting money into play!

“How long” is a controversial topic. Ken French (of Fama and French) has said that 75 years of data may not even be enough to be that confident of your model. Few academic papers that document the efficacy of a system go lower than 30 years.

Personally, I worry about how much faster the world spins each decade. More changes in 10 years now, than in any 50 year period of the 1800s. So I think we can all agree that while studies that go back into the 1800s (usually about the most macro of macro concepts) may be interesting, and that even 50 years is probably overkill.

But less than 30-40 years? I’d certainly be more confident investing in an approach that has performed well throughout 40 years than merely going back to January 2008, which covers one major stock collapse that had its own particular “big story,” as well as a few other lesser events.

ETFers would love to have what Steve mentions for HIS particular needs - the ability to back-test indices as if they were ETFs - a range of indices such as the US and Ex-US global equities/real estate indices, long/intermediate U.S. treasuries, U.S. Credit/High Yield/Aggregate bond indices, a broad commodities index along with the London PM gold fix, and 90-day t bills – that would be amazing.

Of course, if Steve is merely looking for passive optimal asset allocation, it’s more than what is required. I’ll post on that shortly.

But I’d love to see the posts that support short-testing (under 10 years) as being safe, when it comes to the topic of ETFs and asset allocation. I’d be much more comfortable in following a couple of the multi-asset strategies that I’ve been developing (still having some data issues that Paul is investigating, so am stuck for the moment).

Back on the topic of ETFs that go back to 1999…

Denny, thanks very much for that list of ETFs, but are you sure of all those? I count 48 in your list, but I only see 30 in my U.S. database. I checked a couple that didn’t sound right as I scanned the list, SDS and RWM and their dates are 2006 and 2007. I’d guess there are others.

The selection is pretty basic, mostly equity ETFs, until around 2006-2008 (i.e, good coverage of all classes), depending on what’s being tested.

However, one can assemble a decent buy-and-hold diversifier that Steve wants. I’ll post on that topic re what Marc asked above.

All the best,
Ken