Market-Makers' Guesses About Market Direction: Any Good?

by: Peter F. Way, CFA


We know, and have shown repeatedly, that market-makers [MMs] have well-developed senses about specific stocks’ and ETFs’ future price potentials.

But forecasts of major equity market index price moves are far more complicated and difficult for all observers – no one seems to have any reliable, repeating record.

Yet the overall pricing environment is an inescapable part of the MMs' scene. They must have some sense of what can happen; can’t they be of some help?

They do hedge the indexes directly, displaying near-term forecasts and providing warning signs, signs not frequently seen by the investing public.

But is a major market drop starting now? Perhaps, perhaps not.

The principal equity market indexes of interest

Here are the principal equity market indexes, with their symbols as indexes, ETF trackers, and leveraged ETFs. Indexes themselves typically cannot be directly invested in, hence the appeal of the broad-market ETFs. But there are options on most indexes, and market professionals do use those to hedge broad market position exposures.

There are myriad other market indexes, but these are the ones typically given the most attention by investment professionals. The single most "of concern" index for most full-time investors is the S&P 500, with some $17.5 trillion of market capitalizations in its 500 components. The Dow-Jones 30 stocks are all SPX components and make up more than one-fourth of the 500's market cap. Its corporate holdings' size and strength, plus its small number of components make it a focal point for more informally involved investors.

The Nasdaq 100 components have a technology flavor about them, importantly influenced by the biggest-cap stock, Apple, Inc. (NASDAQ:AAPL), but it is accompanied by many other electronic and pharmaceutical issues. The Russell 2000 small-cap index has at times been where the action was because of the capability of many of its issues to be more quick and aggressive in competition with aging industry giants. That advantage never seems to grow old.

The mid-cap indexes have the smallest amounts of capital invested and seem to appeal mainly to individual investors with less than $50 billion committed to the Vanguard product, VO, and to its small-cap brother, VB.

Exchange Traded Fund Index Trackers

Each of these indexes have ETF proxies. It was the inability to trade the indexes directly that brought the ETFs to life originally. A principal advantage of the ETFs tracking the indexes lies in the knowledge that the ETF will closely follow the index's price, not wander off in some "manager's" bright idea of how to "beat the market." That is because market pro arbitrageurs know how to make risk-free profits any time a disparity between the index and its ETF tracker that is more than trivial appears.

The ETFs also have options-market derivatives, which allows us to monitor what the trading in those instruments imply for the future price ranges of the ETFs. Despite the close price movements of the ETFs with the indexes, sometimes the public's involvement with the ETF options clouds the ETF outlook in comparison with that of the MMs.

That is most apparent in the case of the S&P 500 index (SPX). There, the public trades in SPDR S&P 500 (NYSEARCA:SPY) options, while the MMs use SPX options to hedge their broad-market bets. The public is relatively unaware of SPX options, and behaves as though SPY options were more likely to be related to future SPY prices than to coming SPX prices. Perhaps the 10 to 1 price ratio dissuades them, and encourages the MMs. The difference in their impact is evident in the resulting price range forecasts from our analysis of the two underlying issues, one an ETF and the other directly in the index. Here are comparisons of what we see:

(all used with permission)

The pros' appraisals are more critical than the public's, which tends to behave like there should be a more even balance between upside and downside. The past January's data show how and why the difference is important in tactical appraisals of asset allocation between equities and cash or other alternatives.

The appraisal difference is apparent when looking at our Range Index measure. It tells what percentage of the whole forecast range lies below the current market quote at the time of the forecast. For SPY on Thursday (that most right-hand vertical line in its picture), the RI was 47, and the RI based on the SPX was 28. The public saw nearly one half of what might lie ahead to be to the downside, while the pros, instead of being frightened by the goods being on the sale table a week or so ago, saw about one quarter of possible coming events in the negative, and almost three-quarters to be towards ownership profit opportunity.

In the picture of the pros' appraisals of SPX above, back in January the imbalance between up and down possibilities got so pronounced that our picture-making program highlighted it in green, alerting viewers to the buying opportunity. Meanwhile, the public view in SPY plodded on relatively unaware, even though the price changes in the underliers were identical.

Looking at the S&P 500 market index as an equity investment opportunity guide to strategic asset allocation between equities and other asset class alternatives is a common portfolio management approach. The MMs evaluation of SPX prospects has been a strong encouragement over the past two years (and more) to stay heavily weighted in equity assets, despite increasing concerns over rising prices. Here is that longer picture, with daily forecasts taken once a week, over the last two years:

What about emphasis on equity categories, Technology, Small Caps?

The same kind of disparity seen between SPX and SPY exists between intelligent behavior analysis of the Nasdaq 100 index (NDX) and its ETF, the Powershares QQQ Trust (NASDAQ:QQQ). The Range Index for QQQ, seen mostly by the public is now 37, while the MM professionals play the NDX to a Range Index of 8, where it is priced less than 3% of the time.

The history of price behavior (in the index) in the typical 3 months following such forecasts is increases above the price at time of forecast in 7 out of 8 days. The achievement of forecast price range tops (sell targets) has been seen in every past case in the last 4+ years.

Those experiences are not a guarantee of what will happen in the future, but are a bit more encouraging than the QQQ's record. There, the NDX index (being tracked by the QQQ) was indeed higher in 8 out of 10 days (80% of the time vs. the NDX forecasts' score of 87%) and the QQQ forecast sell target was achieved in 94 of every 100 cases, instead of the direct index equivalent of 100 out of 100.

One might well ask why there are any differences in forecast results between things that are driven by the same index, and follow it closely. The answer is that the perceptions of what is likely are far fuzzier for QQQ, with 138 forecasts at RI 37, while at RI 8, NDX has but 28 forecasts - less opportunity to be misled.

The picture comparisons follow:

What is the convincing power of each of these illustrations at the time of late January this year? It ought to be far stronger for the NDX forecasts, which suits the MMs' needs well for shorter time horizons. But has it had usefulness for investors with objectives further out in time? Here is the two-year, weekly look from the MMs' perspective:

The longer-term investor was certainly served well in terms of being encouraged to emphasize technology-oriented equities in his asset allocation decisions during much of this period.

How can the Range Index lead investors to productive, timely investing decisions

We have an analytical tool that allows us to separate subsequent market price movements into groups divided by RIs. That tool looks at all of a stock, ETF, or index's daily forecasts, stratified by Range Indexes, and measures its actual price change out in evenly-measured distances (usually 5 market days at a time) into the future beyond the forecast date.

This tool was created back when computers and their programs were severely limited in several capacities, so it handles only 16 such periods, or the nearly 4 months of 16 weeks.

These measures, from the forecast day out into subsequent market reality, are grouped, row by row, based on their Range Indexes, as indicated in the left-hand column.

The middle blue row shows the average annualized rate of price change for all the forecasts, at each progressively longer time period beyond each forecast date. The rows captioned in green above the blue row measure the progressively more attractive buys as they move away from the average, in Range Index terms, while the red captioned rows below measure the less desirable forecast situations.

The first column of data is a count of the number of buy opportunities at each RI category. The magenta number signifies the current Range Index category of 15, which is less than 29. In this last 4 ½ years to date there have been 414 such forecasts.

This table's content strikingly shows that NDX forecasts below a mid-point Range Index of 40 are followed convincingly by price changes at rates universally above the index's average of 19% a year. Forecasts with RIs above 40 clearly have changes that underperform the average historical rate, actually declining in several cases.

In comparison, price changes following the implied forecasts for QQQ, while distinctly favoring those with RIs below 40, those above 40 are nowhere near as clearly seen to be as underperformers as with the NDX.

The same kind of clear-cut ability to identify favorable market prospects in technology-issue markets also exists in the small-cap arena. Here is the effect of MM forecasts on the RUT index, derived from their price insurance buys directly against the index:

Once again, the MM forecast ability to discriminate likely coming price changes, at least in degree, by means of the balance between upside and downside prospects (the Range Index, above and below a 40 level), is fairly pronounced. What happens when the payoff potential for being right is raised is that an even greater profit payoff appears. Look now at the same table for the ProShares UltraPro Russell 2000 ETF (NYSEARCA:URTY), a 3x leveraged version:

Note that the Range Index for URTY has declined from 34 (2/3rds of the forecast range to the upside, 1/3rd to the downside) to 28, with an even smaller downside history. But more importantly, URTY has in 4 years shown an average rise (during all 1017 days) about double that of the RUT (not triple). Meanwhile, the comparison of annual rates of payoffs between the relevant RI forecast levels of URTY vs. RUT is about 3 to 1.

These black-background tabular displays are measures of simple point-to-point price change averages, raised to an annual rate. The gains could be achieved by simple, buy-hold-sell strategy from the date(s) of forecast to any chosen number of market days later. That would ignore any more advantageous interim opportunities, which when taken would liberate capital for reinvestment in perhaps more advantageous alternatives.

How to improve upon the simple price-change circumstances

A simple sell strategy can both make the capital employed more time-efficient and reduce the risk of loss by the same discipline. What is required is to set a fixed, unchangeable sell price target, and an immutable holding period time limit. We choose to use the top of the forecast price range as the sell target, and 63 market days (3 months) after the forecast date as the holding time limit. Any holding not reaching or exceeding its sell target and being closed out at that point would be timed out and liquidated, regardless of gain or loss resulting.

The sell-target-day liquidation discipline both allows compounding of achieved gains, and forces the reevaluation of alternative investment candidates at that opportunity point in time. Reinvestment in the just-liquidated holding may be the best alternative, but is only to be earned in a timely re-appraisal competition. With a resetting of both new sell price targets and new holding limits, a diversification as to calendar time risks is achieved as well.

Current-day transaction costs are too trivial to be any factor in deterring an up-to-date re-appraisal of alternatives.

When such a discipline is imposed on the major indexes and their various ETFs, here is what history has produced, following prior Range Indexes like today's.

What is in the table

Each row of this table presents (to the left) a price range forecast implied by the self-protecting hedging actions of the MMs, from an algorithmic model that has not changed in over a decade, despite wide-ranging market circumstances. To the right of the Range Index column, the data tell what has happened to each subject's prices during holdings in the time-efficient simple sell-target discipline described above.

The column headed Sample Size counts how many market days had Range Indexes similar to today's, out of the total available of the past 5 years. Proportions of the sample under the discipline described that earned profits are in the Win Odds, the average net gains achieved (including losses) and the average number of market days capital had to be committed form the basis for the annual rates of price return achieved.

During each of those holding periods, the worst price drawdowns from cost were noted and averaged to give a measure of discomfort likely to be encountered on the way to reaching a position closeout, either by reaching a sell target or a holding time limit. The upside price change forecast in the Sell Target Potential as return is compared to the historic worst-case drawdowns as risk in the far-right Reward~Risk ratio column.

The credibility ratio compares the forecast upside prospect with what has historically been achieved as a % Payoff from prior like Range Index forecasts.

Rows in the table's upper half, above the blue averages row, are arranged by the market index concerned, with separate results for each focus on forecasts from analysis of the index, the unleveraged ETF, and the positive and inverse leveraged ETFs.

Those same data rows are rearranged in the table's lower half by the type of instrument being analyzed, for ease of comparison sake.

What the table's data tells us

Is the market of stocks headed up or down from here? Should we run for cover, get flat and be safe, or are there more higher prices for us to tack onto our portfolio's value?

A clue can come from the heightened prospects seen in the leveraged ETF forecasts, the long-positioned ones, vs. the inverse, or shorts. For the big-daddy index of S&P 500, the long-levered ProShares UltraPro S&P 500 (NYSEARCA:UPRO) in prior times of similar Range indexes (52 of them in 4-5 years) saw the disciplined simple sell-target strategy produce further gains 96 out of 100 (actually 50 out of 52). The inverse SPXU evil twin to UPRO spent most of the next 3 months (60 out of 63 market days) being mostly wrong (only 19 out of each 100 experiences making a profit) with pretty trivial 2% gains unable to compete with twice as large % payoffs of +4.5%, earned in only 3 weeks, on average for a better than double-your-money rate.

But wait a minute, why should we be looking at only a +4% gain prospect? Is it because we are late in the market's potential cycle? Yes, the RI on UPRO is 69, meaning only 3/10ths of the forecast is to the upside, and more than twice the gain that can be seen is to the dark side. It may well be that past history has proven UPRO to be a strong momentum play on the upside, with lots of targets reached before worse than -3.9% trouble arrived, but perhaps we can find an alternative that is more comforting than this proposition. Just because something has happened (even often) in the past is no guarantee it will continue to do so again THIS TIME.

The SPXU is looking at a 21% upside, which is likely to only happen with a market in big trouble. Sure, the odds are 5 to 1 against it, but a 20% loss would really hurt. What kind of forecast confirmation can we see in the DJIA, which is a big piece of the S&P 500?

There the win odds are big at 9 out of 10, but the past achievements averaging +3.6% are not what is being forecast at +5%. That produces a credibility ratio among the lowest of any long-position alternative. And the Reward to Risk ratio is pretty crummy for UDOW, with a RI of 54 - more downside than up. So, no, not much "put more capital at risk" encouragement here. Maybe techs will provide some engaging proposition for enlisting the capital liberated by other achieved sell targets.

Yeah, look at this. The NDX keeps coming up daily among the dozen or two best credible wealth-building alternatives, with RIs down under 20 - a 4 to 1 gain proposition that has come home at a rate of 97 out of 100 times, or only 2 losses out of 76. It looks good with an upside sell target more than 6 times its worst-case average drawdown experiences among such prior adventures. But we can't easily do it with the index options, which means setting up a margin account and learning enough about options to have the broker comfortable that you're not going to have your lawyer cry all over his expense sheet if you might lose any money.

So what does the ProShares UltraPro QQQ (NASDAQ:TQQQ) offer in its place? Well, 86% historic win odds is about 7 out of every 8, and the net gains have been at +10%, even including that 8th one. Getting them home in 7-8 weeks puts the annual rate up at +90% if history is repeated, and a closeout ought to be in hand before the summer starts. Lots of good, credible encouragement here, and there's nothing scary and convincing from the inverse SQQQ side.

Can the small-cap world be any more appealing than the tech world? Well, maybe. RUT and related ETFs are in the 9 out of 10 wins odds, with historic net gain % Payoffs better than their forecast sell target returns. And URTY is offering double-digit upside from a RI of 28. Past achievements of even more in only 6+ weeks seem like a TGTBT (to-good-to-be-true) goal, but it's been done over 50 times in the last 4 years. Perhaps we need to do some more intensive DD on the small-cap front. Its inverse offering's history is truly ugly, so that's good on the "put idle capital to work" front.

How about mid-caps? They don't look bad as a generic proposition, just small returns in scale, earned in 4-5 weeks in the past. But comparisons with techs and small-cap arenas suggest that with the rest of the market where it is, it may be best to let the mids develop until better prices and prospects get stronger market support from everyday portfolio adjustment activity by the big-money fund clients.


For big parts of the equity market, including much of the big-cap generals and diamonds, there's no good reason to be fearful - nor wildly optimistic. The market as seen by SPY and DIA may just stumble along for a while.

But attention seems to be deserved among the techs and small caps, so put on your Sherlock deerstalker hats and search out some winners. The environments there apparently are constructive, with expanding commitments firming things up.

Or is all this stuff just an elaborate smokescreen of details devoid of any fundamental basis when the "real" analysis says things are already expensive?

Just remember, folks with 7-8 figure a year compensation jobs have to prove their worth every month, every quarter, every year, by making the right bets with the firm's capital. That's where this data comes from - the way the smart (and aggressive) money protects itself, continually. A few mistakes may get made, and hungry young tigers lie in wait to step into any serious vacancy, determined to make their own marks and early successful retirements.

No guarantees, but the past record continues to be impressive, and we stay on top of it.

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.