Seeking Alpha
, Blockdesk (2,292 clicks)
ETF investing, CFA, portfolio strategy, long/short equity
Profile| Send Message|
( followers)  


  • Our recent article on gold stocks concluded that market-makers did not have any special pricing insights into the group – nor did many others, making the group volatile and unpredictable.
  • That belief was based on an impression of light volume traffic in these stocks by most institutional clients.
  • That may have been a bad presumption for at least one of the ETFs, whose daily average trade load of $1½ billion, renews its entire $36 billion AUM each month.
  • That so, it deserves a more careful Intelligent Behavior Analysis of market-maker judgments of its likely coming prices.

The mother lode

Our recent article on gold stocks concluded that market-makers did not have any special pricing insights into the group - nor did many others, making the group volatile and unpredictable.

The biggest ETF capital collection of gold-based commitments is SPDR Gold Shares (NYSEARCA:GLD), a $36+ billion hoard that makes up more than the total capital placed in all 22 other major precious metals ETFs.

With a share price around $125, a round lot of GLD is no foolin' around for most individual investors. The recent average daily traffic in it, according to Yahoo finance, is nearly 7 million shares a day or a tab of over $1.5 billion (with a b).

Some big piece of that activity has to be handled in ugly lumps called blocks on the street. The very serious poker table regularly in play is no T-V championship event, but an every-day knock-down, drag out fight over who gets what. All the traders there have up-to-the-minute data on whose mine is flooding, how much metal the electronic scrap recovery lines are turning out today, where the underground-worker natives are restless or the above-ground ones are kidnapping people, and what deals Vladimir may be posing with the Chinese - or at least what is most likely.

But more importantly, they know well the big-money-fund clients their market-making firm is fronting for on the multi-thousand share GLD trade that is in the spotlight right now. And the next one, and the one after that...

Everyone has his/her idea of what today's gold price should wind up at. And tomorrow's and next month's. As Noccolo Machiavelli said, "Everyone has his reasons" but as always, it's the guy with the biggest money muscle - of the moment - that sets the price now.

The "now" price is the closure of someone's strategy started a while ago. The strategies still open, and those starting up, or poised to be started, are at the mindfront of the Market-Makers [MMs] in money-mortal combat with one another. Expectations are paramount, and none at the table are careless with their anticipations.

Here comes the "tell"

All of these pros are very rational, skilled and experienced folks. They know when to dance around the edges. But they won't put the firm's capital at any more risk of loss than can be stood by their own personal reputations (and their big-pay jobs). So they know 1.) when to buy adverse-price protection, 2.) what they can afford to pay for it, and 3.) how best to structure the armor so they and the client get the best deal possible out of it.

Those actions are what tells the knowledgeable behavioral analyst just how far the negotiators figure that prices might get pushed in the coming days, weeks, months. Not because the MMs are going to stick around for more than several hours, at most. But the pieces of their protection may (often do) have contractual lives of up to months, and the parties involved in those contracts have the privilege of seeing that their rights are upheld whenever it is appropriate, whether immediately or just before expiration. And that is what the protection's price contemplates when it is bought.

So the MMs' self-protective instincts produce price-range forecasts for everything that they deal in, including shares of GLD. Just as they have, daily, for years. We thought it might be of interest for you to see how those forecasts related to market prices when we go back to mid-2008, before things got messy in the fourth quarter of that year, and thereafter:

(click to enlarge)

In this picture, the vertical lines plot the ranges of price for GLD believed possible in coming days and weeks, based on the protections bought by MMs in their hedging activities. The yellow dots are the actual GLD closing price on the day of the forecasts.

Over the years, we have seen that the imbalance between upside and downside price prospects, set by the fulcrum of the closing price on the expectations range, has predictive content for the subject's market quote in coming times. We take a measure, referred to as the Range Index [RI], of what part of the whole possible price range lies below the contemporary market price. It is shown through the pictured period as the red line, with its scale on the right.

To illustrate the relationship between the RI and subsequent price changes, we have put together an analysis tool that looks at a subject's price changes in one-week (5 market-day) intervals for the next 16 weeks, or almost 4 months. It looks at every market day, and stratifies (according to the Range Index value of the beginning day) the price changes from that forecast day to each of the progressively longer (potential holding) periods.

The results table for GLD displays what those price change averages were, when cumulated, progressively row by row to the highest RI from the lowest-value one. If you check the picture above, it should be apparent from the red line that the RI for GLD never got above 50, so all of the 1130 days' results are in the blue row, as an average. (When subjects do have RIs above the mid-point, they get cumulated upwards from the table bottom rows, again to the middle, where all are shown.)

(click to enlarge)

A couple of things should be apparent from this table.

At lower Range Indexes (larger Reward : Risk ratios), there are very attractive annual rates of price gain in GLD, compared to its blue-row average. Those gain opportunities persist out in time to the three-month (65 day) period.

The column labeled #BUYS counts how many instances of such forecasts meet each row's minimum reward-to-risk description at left.

So good, big opportunities are not all that common, with 15 : 1 chances (RI=< 6 or 7) occurring only about 3% of the time (38 / 1130). The significance of the magenta 267 is that this 5 : 1 row is where the current RI is found. With not now attractive results.

Moral: It's important when to pick your fights with GLD. But there can be very rewarding times to do it, as is the case with most equity investments. Let's see how the opportunities can be increased and/or the likelihood of good payoffs can be increased.

What are the odds of being right, and how big can the payoff be?

Let's extend our look at GLD's RIs back to mid-2008, before the mortgage issues hit the fan, and understand the price effects more completely. Here is a scatterplot relationship of all of GLD's forecast Range Indexes since mid-2008, over 1500 of them, running through the troubled period (including March 2009) to the present, compared with the best price gain opportunities present in the next 3 months.

(click to enlarge)

You're right, that doesn't look like a very tight fit between RI and Profit Opportunity for GLD in most cases. Pretty poor, in fact, since the biggest 3-month gain opportunities appeared among RIs of 20 to 40 with little to no concentration. And in a RI span of -30 to +50, RIs of 20 to 40 are pretty high.

How can this be so, when the travel of prices following Range Index levels in the prior table looked so orderly? The answer is that these maximum gain opportunities were found by simply looking for the largest price in the next 3 months, regardless of how soon or how irregularly it arrived. It only tells us what we already suspected, that the market is a noisy place.

That only begs the question of how soon to get off of each ride. We each have only so much capital to invest, and just so much time to do it in. Why not adopt a standard procedure to be followed with every investment and see how well it works? To save you the time and anxiety, here is one that we have found that works quite well, once you get used to what it does.

Decide on a maximum Range Index for your buys. From the forecast, use the top of the price range as a sell target. Watch each day's closing price. The first day the target is reached, sell the stock or ETF on the next day. If that hasn't happened in 3 months (63 market days) sell it regardless, and seek a new investment with the proceeds.

Between the buy and the sale the investment's price may go below your cost. This risk period is when you earn your investing reward, as the investment usually will recover and go on to a profit because the MMs' forecast has the power of experience and good information. If it never goes down and you hit the target quickly, you were lucky; next time likely will be the more usual. If the price never recovers to your buy cost, that is the inevitable infrequent outcome of playing a serious game against other skilled or lucky contestants.

The MMs are good, but not God. They are not always right, because sometimes situations change without warning, or things that are generally regarded as improbable occur. But their profit batting average (ODDS of winning) is far better than average, and becomes your advantage.

We test our Market-Maker information constantly using this strategy to find out which subjects prove to have the best Win Odds combined with substantial profit payoffs, net of the inevitable few 63-day losses. Over the 1500+ day life of GLD since mid-2008 seen in the scatterplot above, here is how the MMs have scored:

Our first quality-control check on tables like these is the column headed "% of Winners." This one is pretty ugly, with the best result coming from one RI level (of zero, where the entire forecast range is above the current price) that had only 6 examples, a very small sample. The two next best show fewer than 3 out of 4 forecasts as producing profits.

Then let's check the Average % Gain column for realized payoffs. Many of these show up with net losses, not gains. Ugh! How do those results compare with the upside targets being hedged against? There is practically no justification or credibility to the upside claims.

The only conclusion we can come to is that the MMs approach volume trades in GLD as being near impossible to foresee. So they build very wide protection areas into their hedges and dispense with much of any concern about making profits on the hedge beyond covering any losses that needed to be taken to fill the trades. The fact that no Range Indexes exist above 50 is a tell-tale that the upper half of their forecast ranges is fictional.

Elsewhere they can use their true privileged knowledge and expertise to make substantial profits, so they deal with GLD trades as quickly and simply as possible in my opinion.

We come to this conclusion because a very different picture appears where the MMs have continual exposure and repeated demands for their trading expertise. Look at ProShares UltraPro Dow 30 (NYSEARCA:UDOW) as an example.

UDOW is a 3x leveraged long ETF holding the 30 Dow Jones Mega-cap stocks in which virtually all big-money funds have serious investments. All of the skills and resources held most essential by MM firms are involved here, so it figures they should score well on our test. And they do:

Here, note the % of winners among Range Indexes up to 45. Also see how evenly spread out among RI levels are the number of Buys (forecasts). That is because there is a great deal more very careful thinking and analysis going on, making distinctions much finer. That is also evidenced in the column of Upside % Target, which gradually declines as RIs rise.

The credibility of those Upside targets is well supported by the achieved results in the Average % Gain column.

An acid-test comes in the Days Held column telling how long it typically took a forecast to reach its sell target. There are no over-40, 2-month holding periods here. Combined with the gains, triple-digit annual rates of price-gain returns are prominent. The only loss category is at Range Indexes with 90% to the downside.


It is evident that the MMs can find some times that are opportune to enter a capital position in GLD, but unfortunately, this is not a good time, in their evaluation, given many other choices.

We are not misguided that Market-Makers have little special insight into the wild free-for-all of "investing" in precious metals. They try to do their jobs as well as can be done, but the complexity of the world-wide audience make-up, and the diversity of focus may tend to keep the subject matter far from rational resolution except occasionally after the fact.

At that point everyone knows what should have been done and half the buyer-seller audience has probably done it well. For the others, "just wait for the next move."

But what wealth-building investors can afford to spend their time that way?

Alternative choices of ETF investments, like the illustration of UDOW, put the MMs' skills and resources to far better, time-efficient commitments of capital. But at its present Range Index of 67, a small upside, and poor past gains, this does not appear a best-time choice for it, either. A patient, brief holding of cash may clarify opportunities, as market prices continue to percolate.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Source: SPDR Gold Shares - A More Intensive Pricing Study By Market-Makers