ariston跨市场资产策略笔记AUG 27, 2026Edited

Technical Analysis: The Basics and the Not-So-Basics The Basics and the Not-So-Basics of Technical Analysis

Rethinking Trends, Fractals, and a Multi-Timeframe Perspective

_____________________________________________________

27 August 2026

Notice: By reading, subscribing to, browsing, forwarding, saving, or using the content of this planet in any way, you are deemed to have fully read, understood, and agreed to this disclaimer;If you do not agree, please stop using it and leave immediately.All content on this planet is solely for personal research, information organization, and learning and discussion. It does not constitute any investment advice, buy or sell recommendation, solicitation of an offer, promise of returns, asset management service, or investment advisory service, nor does reading, subscribing, interacting, or using it create any legal relationship such as client, agency, trust, or advisory relationship. The content does not take into account any recipient’s financial circumstances, investment objectives, risk tolerance, or suitability requirements. Recipients should verify the information themselves, make independent assessments and decisions, and bear the investment risks and consequences themselves. The information herein may come from public sources, third-party data, market information, personal judgment, or model analysis, and may be delayed, incomplete, erroneous, biased, or invalid; the author does not guarantee its authenticity, accuracy, completeness, timeliness, or suitability. Past performance, historical examples, backtesting results, or successful calls do not indicate future performance and do not constitute a guarantee of returns.Without permission, you may not copy, screenshot, screen-record, forward, adapt, compile, or use the content for purposes such as stock recommendations, fundraising, managing money on behalf of others, or commercial promotion; all consequences arising from such actions shall be borne by the persons responsible!

Subscription to the FICC&MACRO main column:https://www.meridianresearch.app/home?product=cross-market-strategies

CNData, investment research notes and commentary subscriptions:https://www.meridianresearch.app/home?product=asia-strategies

USData, investment research notes and commentary subscriptions:https://www.meridianresearch.app/home?product=us-strategies

Additional research on commodity markets:https://www.meridianresearch.app/home?product=us-strategies

Related link: Sean’s global market technical analysis column:https://www.meridianresearch.app/home?product=vega-trading

I. Rethinking Trends, Fractals, and a Multi-Timeframe Perspective

Markets are chaotic and unpredictable by nature, yet they still form self-similar recurring patterns. These patterns, as the results of decisions, in turn influence participants’ decisions to a considerable extent By reading price action, market participants can learn how broad pricing behavior unfolds. This does not mean that simple technical analysis can produce astrological-style predictions. On the contrary, compared with prediction, reading price action is more about identifying and understanding “what is happening.” To identify it well, the first requirement is to have a market language for interpreting price action。

Chart:NDX(Monthly chart, as of 2608)

As a fundamental building block of market language, one perspective is the fractal perspective.The concept of fractals originated in mathematics and was later widely applied in other fields, including audiovisual media, physics, and biology. The advent of fractal geometry made it possible to measure and describe natural structures once considered immeasurable, including, but not limited to, clouds, coastlines, protein structures, snowflakes, and trees.

The key features of fractals are iteration and self-similarity. The best-known iterative model is the Fibonacci sequence, which was later also applied independently to chart pattern recognition in technical analysis. Those of you who have used it to measure price targets have, in fact, already encountered the philosophy of fractals without realizing it.

The feature of self-similarity is linked to the more fundamental multi-timeframe structure of financial markets.Countless participants and pools of capital engage in the market across different timeframes, based on their own decision-making rationales and horizons. Most sovereign entities and large financial institutions make decisions on weekly and monthly timeframes, individual traders tend to focus on the daily timeframe, while many algorithmic trading strategies operate down to the minute or even the second, providing liquidity.

Figure: The world of patterns arises from visual cues at every moment

A foundational understanding of technical analysis from a fractal perspective is that, since individual and collective behavior is underpinned by the same fundamental human nature, or by the algorithms derived from it, seemingly irregular price movements still display highly self-similar recurring patterns across multiple timeframes.

In other words, why are the two features of fractals—iteration and self-similarity—so important? Because together they form the fundamental assumptions underlying the entire system of technical analysis: 1) Market pricing is the full outcome of the iterative interaction of collective behavior; all information is already reflected in prices before it appears before people in the form of language or financial data. 2) Collective behavior has similar characteristics, so judgments based on today's information can yield a specific expected distribution of future outcomes. 3) The distribution of outcomes holds across multiple timeframes. This means the method can produce consistent results on a particular timeframe, as well as across iterative timeframe structures in both directions: seconds to minutes, minutes to hours, and hours to days, weeks, and months.

Once we understand fractals, it is natural to ask why they are the first lesson in this handbook. Fractals are the most fundamental building blocks of any pattern or trend. Whatever timeframe you use to make decisions, execute trades, and set recognition criteria, fractals are the most basic units of construction; this section explains their mathematical definition.

Note that this article discusses only local extrema fractals formed by candlesticks in technical analysis; these are not the same as self-similar sets in the strict sense of fractal geometry. The definition uses only the high H and low L;the open O and close C are used only to draw the candlestick body and for subjective early confirmation。

Let a finite K-line sequence consist of a total of T completed bars; each bar contains four fields: the open, high, low, and close. The notation is standardized as follows:

Given a positive integer radius, take the same number of bars to the left and right of the t-th bar as the center,

forming a symmetric window of odd length:

A top fractal and a bottom fractal are regions in which the center bar forms a high or low within the window:

The indicator function is defined as follows; the strict inequalities ensure that the central extremum is unique within the window:

Here, k denotes the radius of the left and right windows; the indicator function takes the value one when the event holds. The fractal is attributed to the central candlestick, and standard confirmation occurs when the final candlestick on the right is complete; the same center may simultaneously qualify as both a top and a bottom fractal.

The chart above shows a standard three-candlestick fractal example, in fact, compared with the three-candlestick fractal, I would prefer to call it a one-delay fractal; this name more clearly conveys the essence of the fractal: it is a fractal pattern that requires at least one candle of delay for confirmation, and traders need to confirm their execution entry near the close of the candle after the top or bottom candle.

The criteria are as follows:

The fractal’s shape is attributed to the central candlestick, but it cannot be confirmed until the candle after the center has closed. Therefore, the earliest a closing decision can be made is after that right-side candle closes; if execution is at the next open, it must be deferred by one more candlestick. In practice, if you already have a directional view of the trade as a whole, it is better to act at the close, or even upon an intraday breakout, so as to enter as early as possible relative to others.Ultimately, your trading profits come from other people’s losses. Given that premise, how to get ahead while maintaining a certain degree of certainty is a key objective in applying technical analysis. It is a form of arbitrage in which the certainty for you is greater than the market average.

From day one, you should know that, excess returns come only from:

arbitraging under an advantageous structure (relative to the market average).

The chart above shows a fractal structure with K=2, which is more widely used by swing traders; I call it a two-delay fractal. The central high of a top fractal is higher than the highs of the four comparison candlesticks on either side, while the central low of a bottom fractal is lower than the lows of the four comparison candlesticks on either side.

In actual market movements, there can also be cases of tied highs or tied lows. Unless the equality rule is specified in advance, the same data may produce different results on different platforms or in different code.For equal values, we accept that they simultaneously carry the dual attribute; that is, within a higher-timeframe structure, they can serve as both a top fractal and a bottom fractal to confirm the structure.

Besides confirmation at the close, there is a more aggressive form of confirmation based on an intraday break through the central candle’s body. This pattern often also appears as an engulfing pattern near a key price level on a higher time frame, and confirmation is often stronger when the two occur together than when either occurs alone, with no gap being preferable here, since both sides of the market can fully contend at this level.

Definition of the upper and lower boundaries of the central candlestick’s body:

Record the index of the first subsequent candlestick that meets the breakout condition as the breakout confirmation time:

Breakout confirmation indicates the direction and strength of a reversal, but while the k candlesticks on the right have not all completed, it remains an early confirmation of a candidate structure and does not mean that a standard fractal has been definitively established. Under the same strict criteria, a five-candlestick fractal imposes a stronger local-extremum condition. Every five-candlestick top fractal necessarily qualifies as a three-candlestick top fractal, and every five-candlestick bottom fractal necessarily qualifies as a three-candlestick bottom fractal; the converse does not hold. This distinction will be explored in detail later; in fact, Choosing which one to use is a trade-off between sensitivity and confirmation.Which one to use depends on multiple factors across different multi-timeframe scenarios and the bigger picture.

Beyond that, actual markets are often not as mathematically perfect as a model; more candlesticks may be needed to complete the final confirmation. However, confirmation is objectively defined and can be extended, so this is where flexibility is needed.

Figure: A bottom fractal reversal in the U.S. Dollar Index

From the very first day you enter the market, you need to make yourself adopt and stick to a multi-timeframe (MTF) perspective. This point wasAri’s WEEKLY WRAP 260823》thoroughly explained in:

More important than structure is a multi-timeframe (MTF) perspective. The Buddha sees eighty-four thousand insects in a bowl of water; the prices you see with your naked eye are the same. Behind every one or several bars lie dozens, hundreds, or even thousands more bars. Each bar at each level is linked to different market participants and algorithmic models. This is the essence of multiple timeframes...

The essence of a top fractal is a process of price acceptance, price rejection, and price acceptance on a smaller timeframe.The bulls’ advance loses momentum after making a new high, while the bears complete an engulfing move after the new high, first negating the higher bottom fractal and higher top fractal on the smaller timeframe. Next, consensus traders follow closely by confirming acceptance of a lower bottom fractal and a lower top fractal on the smaller timeframe, which is reflected on the larger timeframe as a break through the body of the center candlestick.

Figure: The U.S. Dollar Index on the daily chart

Figure: The U.S. Dollar Index on the minute chart; what happened inside the red circle behind that single candlestick

The process of improving your market feel and eye through diligent practice is the process of learning to see deeper at a glance.Many people look at a candlestick and see only its rise or fall. But an exceptional trader can see one, two, or even three layers of fractal iteration behind it, as well as the further battles between bulls and bears taking place there.

Always pay attention to what is happening outside the timeframe you are betting on. What may seem unrelated to you is actually closely connected. The contests on larger timeframes form the backdrop to your bet, while those on smaller timeframes construct your timeframe at the very margin, Your timeframe is the result of the iteration of larger and smaller timeframes.

At this point, we can build a basic method for identifying trends and non-trending conditions. Though basic, in practical terms it can outperform the 99% trend-identification systems you see on the market.Everything in the manual is practical, with the aim of making it possible to put what you learn to use immediately. Only by applying it in the actual market can you continually improve how well the techniques mesh with your own system, in service of the goal of earning profits in the market.

Generally speaking, because one-bar delayed fractals appear frequently in price action and create too much noise, a trend-identification system should be built on at least two-bar delayed fractals. The use of one-bar delayed fractals will be discussed in a later section of the manual, including how to retain the most sensitive leading edge while also achieving as much stability as possible, in order to obtain better probability, payoff, and expected outcomes.

1) Uptrend: higher bottom fractals and higher top fractals

2) Downtrend: lower bottom fractals and lower top fractals

3) Ranging price action: all other cases

Chart:NDX(Monthly chart, 1990-2004)

The chart above is the monthly chart of NDX in the 90s; the lower panel shows statistics on price changes within trends. The maximum gain within a trend reached 600%. If you can clearly identify a trend’s persistence, and were trading in the market in the 90s, you would not have missed this major rally either.There is no need to worry about supposedly missing out, because the market will tell you itself when it is coming to an end.

As you can see in the chart, a series of higher lows and higher highs formed an uptrend, shown by the green background, while around the 2000 year and -2001 year a break below the latest bottom fractal at the time (a variant of the immediate-use approach, in which a renewed rejection of the previous latest bottom or top fractal can also be regarded as the end of a trend or a trend reversal)announced the end of this nearly decade-long advance. The bear market began and persisted until 04, when it finally eased. If you had used so-called upward trendlines to mark support at the time, you would have lagged the fractal confirmation by at least 4 weeks. If you had used a medium-term moving average, you would have lagged the fractal confirmation by at least 6 weeks.

Chart:MU(Daily chart, as of 260827)

The chart above shows price action from the beginning of MU to date. It is clear that the two major trends are both identified very cleanly through the formation of fractals, capturing swings of 60% and 165%, respectively, and retaining profits of 53% and 140% when the trends ended. If you traded only these two moves, the compounded return would be approximately 370%, while MU’s total gain from the first trend at the beginning of the year to date is approximately 270%.

Any tool must be used with the strictest scrutiny. To borrow Munger’s words, what you always need to know is not what you can do, but what you cannot do—where, exactly, you will die. If you know where you will die, never go there. Knowing what not to do is often far more important than knowing what to do.。

The problem with fractals is that they only consider local structure, and do not inherently account for the broader context or MTF structure.To address this, combining rectangles with a broader time-frame perspective can lead to better decisions:

Chart: MU (daily chart, as of 260827)

Taking MU as an example again, most trendless periods and losing trades occur within rectangle structures. If you can identify these structures, when to act—and when it is appropriate not to act—becomes much clearer. To some extent, this also lets you intelligently avoid the weaknesses of fractals.

Rectangles can likewise be identified using fractals, including the threshold method and the extreme-line method. There are actually even better ways to identify them, but let us set those aside for now. In the chapter on using fractals, we will first develop a thorough understanding of them; later, we will gradually expand on the discussion and help you find the right tool—the one that plays to your strengths—for every action.When participating in the market, discard the traditional idea of compensating for weaknesses. Trying to shore up weaknesses is useless; we should make the best use of everything and use the best available tool in each situation.

...

First, the threshold method:

After measuring price changes within fractal trends (shown in the lower panel of the chart), for each instrument we can derive a consolidation threshold line—that is, an indication of the approximate range in which the rectangle characteristics of a particular instrument lie. For MU, for example, the threshold line is approximately within 20%. This is a better way to identify ranges and trends than traditional channels and indicators such as ADX.

Every instrument has its own set of characteristic values within a given time frame. These values are determined by the instrument’s main participants during that time frame. Traditionally, some call these main participants the “market makers”; Graham’s term, “Mr. Market,” can also be used. Some instruments are made up of speculative capital, some are major allocations for insurance funds, and others are linked to broader cross-market instruments. Behind them are different market figures—Mr. A, Mr. B—and particular pools of capital are often linked to particular behaviors. Price suppression, accumulation, and distribution are all deliberately or unintentionally kept within the threshold, and the trading range within that threshold naturally forms a dynamic rectangle. Sometimes this range departs from the rectangle that the eye can directly see and appears in various forms. But you can see through appearances to the underlying reality: behind all these varied shapes, they are all rectangles.

The stage approach represented by Stan Weinstein and the price-volume approach represented by Richard Wyckoff both attempt to identify and classify consolidation ranges. The former divides them into Stage1 and Stage3, while the latter divides them into accumulation and distribution. In essence, whether they are called 1 and 3 or accumulation and distribution, non-trending ranges are the enemy of all traders and investors. Ultimately, everything comes down to trying to do one thing: act only when it is time to act, as much as possible.

Both of these approaches have some fundamental problems, which is why I basically don't use them. In fact, if you use them directly, they both have negative expected returns and will lose money over the long run. Later, I’ll share how I’ve adapted the material on these approaches.

Chart: CMX gold (daily chart, as of 260827), rectangles identified by the feature-threshold method (areas to avoid)

...

Besides the threshold method, there is also the extreme-value method.The extreme-value method essentially treats trend-ending fractals that reject trend continuation as extreme highs or lows, then adds buffer zones above and below these points. Until subsequent price action breaks out of a buffer zone, the area can be viewed as a directionless range in which price continues to be rejected at that point:

Chart: Changfei Optical Fiber (weekly chart, as of 260827), rectangles identified by the extreme-value method (areas to avoid)

Taking the chart above as an example, after the downward fractal swing in year 21 was invalidated, it effectively marked an extreme price at the bottom of a rectangle. This price matters because bears were rejected there. After the upward fractal swing in the middle of year 22 was capped, it effectively marked an extreme price at the top of a rectangle, and this price matters because bulls were rejected there. These rejections at the top and bottom define a value range. Until price breaks out of that range, mean-reversion trading will have the upper hand and trend trading will take a temporary pause. The larger the swing within the trend before the rejection, the stronger the rejection price.

The material above Section Six is written for swing and position traders. From here, Section Six is for short-term traders who scalp. First, an important disclaimer: I do not want any of you to become short-term traders, because short-term price fluctuations offer no excess returns. There are no real profits in scalping; the money you make here is more like a wage or compensation. The only way out of this kind of trading is to use leverage, but that also tends to be a dead end.

For the swing traders I hope you will become, daily P&L fluctuations and intraday market structure are things that need to be blocked out completely. Repeated exposure to noise will make you give up all your favorable odds—you’ll cut off your right tail, take on all the stop-loss risk, and then voluntarily truncate the potential returns that go with it. Add normal probabilities to the mix, and the result won’t be good. This is also where the vast majority of retail traders operate, which is why they lose money, or make money occasionally but lose over the long run, or earn less than they would from a wage. In short, even if you use it, I don’t recommend making it your main tool, because it really isn’t a good habit. It’s a bit like drug use: it can damage your brain and your nervous system’s habits to the point that you can no longer truly ride a trend. It’s an addictive behavior, and its physiological effects are fairly similar to those of gambling.

Chart:SMH(Daily chart)

Chart: SI (daily chart)

If you look carefully at the image above, you’ll get a very clear answer: 1)A truly smooth rise will not produce a top fractal, and a truly smooth decline will not produce a bottom fractal; once a fractal opposing the trend appears, 2), the vast majority of cases will enter an ambiguous zone of consolidation, stagnation, or reduced volatility. This does not mean that the trend over a longer time frame has come to a halt, but it does mean that the best window for deploying capital has passed.If your goal is to maximize capital utilization through scalping then what you should be looking for is a definable, smooth rise or smooth decline.

The best approach here is also to use a MTF perspective. That is, if your holding period is measured in hours, look for a smooth move on the daily chart. If your holding period is 15 minutes, work to identify a potential smooth move following a breakout near a key price level on the hourly chart.

This is also an if...then logic: if a rise or decline is smooth, there will be no pause before the initial pricing is complete. Initial pricing is often the climactic stage of expanding volatility. Many traders have discussed this, for example EP and others; they are all, in fact, betting on the initial pricing.

You’re all very kind to think so highly of me. I’m not important, and I have no intention of becoming famous. I prefer to keep a low profile; the content in our private group is for our members, and we try not to share it externally.

Thank you all. I’m grateful to all my friends.

...

(End of full text)

To be used alongside the US session, A-shares, Hong Kong stocks, and the commodities group, Only decide when there’s a real need.I know everyone supports it, but don’t waste money or spend it recklessly.

U.S. markets:https://www.meridianresearch.app/home?product=us-strategies

A-shares/Hong Kong:https://www.meridianresearch.app/home?product=asia-strategies

Commodities: https://www.meridianresearch.app/home?product=commodity-strategies

Note: All content in this document is for reference only. It does not constitute any investment advice, nor does it serve as a basis for any laws, regulations, or regulatory policies. Investors should not use or rely on such information as the basis for decision-making.

Note: All content shared within the group is for reference only. Recipients are not deemed to be our clients by virtue of reading the group’s content, nor does it constitute an invitation to anyone to buy or sell securities or other investment products, or a basis for trading. Markets involve risk, and investing requires caution. All information and views in this group do not constitute investment advice. Past performance does not represent future results. Investors should carefully assess their own risk tolerance, make independent decisions, and assume full responsibility. Under no circumstances shall we be liable for any losses arising from the use of the group’s content. The relevant data may be delayed or inaccurate, and we accept no legal liability for the completeness or accuracy of the information.

Sean Vega Trading2026/10/3 16:21

Michael Burry added to his AI shorts on September 22, I said these names had further to run—and that he would need another post explaining why he remained short. I posted this on the same day.

On September 28, he doubled down, shifting shorts into puts on MU1,074.89, NBIS242.81, CAT845.42, SOXX588.90 and NVDA233.95, increasing his PLTR188.75 and NDX30,807.93 put positions, and opening MET puts.

Here is how the nine names performed through the October 2 close:

And on October 2 he was back defending his thesis. Right on cue.

Sean Vega Trading2026/10/29 13:29

Yesterday's FOMC proved to be a non-event, as we expected, especially compared with the previous two meetings. NDX29,828.29's intraday range was just 0.78%, versus 2.81% on July 29 and 1.67% on September 16.

AMZN247.13, INTC92.90 and SNDK1,380.12 report after the close today. So far, all the major tech earnings have been met with negative reactions regardless of the numbers reported, except MSFT, whose gains are already fading in premarket.

Despite the NDX29,828.29 selloff over the past two to three weeks, SPX7,748.03 remains only about 1.5% below its ATH. This is consistent with my call for a bond rebound that would support equities in the ST. As I have mentioned several times in previous Tactical Positioning, I do not expect a large selloff into the election. I think the market is underpricing the risk of a large selloff after it. At the same time, there is a small chance of an explosive post-election rally, as we saw in 2016, though I put very low odds on that outcome.

Months ago, when oil rebounded from our pivot levels around 70 into triple digits, I outlined the path into the election. The broad framework was the 70s on the downside and 100-120 on the upside, with 100 as the normal upper bound and 120 if the situation got out of control. That has largely played out.

Once again, there has been no progress whatsoever toward a peace deal. Traffic through Hormuz has improved significantly, but as I have said, keeping oil prices high is in Tehran's best interests. It does not need an explosive spike. As long as oil stays elevated, time is on Tehran's side, not Washington's.

Oil inflation generates second-order effects that can make inflation very sticky. We are now eight months into the war, and those effects have yet to materialize. I expect them to emerge next year.

The bond rally is technical rather than fundamental. I called the yield tops weeks ago, and Treasuries have since rallied hard across the curve despite no visible drop in oil prices. I expect this rebound to end in the coming weeks as the positive correlation between oil and yields reasserts itself. Unless we see a material change in the Middle East, rising yields across the curve seem to be the only path forward.

This technical bounce has been among the most lucrative trades this year. Congrats to those who took it. As always, trail your stops to protect your gains.

NQ1!29,625.25 was rejected at the horizontal resistance following the earnings releases. I would have shorted that rejection had it happened during RTH. ES1!7,748.50as been trapped in a very narrow range between 7,660 and 7,840, which I expect to break as soon as Election Day.

AMZN247.13 continues to trade within our LT rising channel. ST resistance sits around 290, with an extension toward 302, while ST support is around 225. I doubt tonight's earnings will take it to either end of that range, whether in after-hours trading or by tomorrow's close.

INTC92.90's generational channels capped its upside a few months ago. Given the cyclicality of the industry and where we are in the broader cycle, the odds favor tests of support at 67 and 52 in the coming months. That is consistent with our expectation of a selloff after the election.

I have received a lot of feedback over the past few weeks, and based on that feedback, there will be major changes to the content I provide to this community.

Tactical Positioning will cease, as the majority of you fail to appreciate the finesse behind these charts and their predictive power. That does not mean all the content from Tactical Positioning is going away.

Future content will focus primarily on trades, setups and the analysis behind them. Subscribers who were here in 2024 and 2025 will find this familiar. This is a return to the classics. Shorting near the top ahead of the tariffs in Q1 2025, going long near the bottom and shorting gold near the top are just a few examples.

Charts and analysis from Tactical Positioning that are relevant to a trade or setup will continue to be provided. Updates will come when I think they are needed, rather than on a weekly schedule that can be untimely and contain more irrelevant information than relevant information. This post is an example of what future content will look like.

I also want to align expectations. I do not like regularly scheduled updates because sometimes there is simply nothing to update. No significant price action is expected, and the best thing to do is step away from your screen and enjoy your life.

When significant moves are happening, I could publish several posts in a single trading day, as I did ahead of Liberation Day and during the first few days of the Iran war. That is how it should be.

G-Test Northstar Macro2026/10/2 10:26

10. Leveraged or inverse ETF

G-Test | Simulated examples only. Research / Execution Review 20261002. No broker transaction; no Sean or Vega trade.

Design-boundary example, no trade inserted. NDX research could be expressed through QQQ, TQQQ, or SQQQ, but a long inverse ETF is not an NDX short fill.

Keep the trade side separate from the research direction. Record actual ETF prices and shares; do not multiply the index return by a leverage factor across a holding period. The reviewed catalog lacks these ETF entries; separate instrument setup is required.

Research reference: NDX. This reference is not an execution.

G-Test Northstar Macro2026/10/20 0:41

Technical Analysis: The Basics and the Not-So-Basics The Basics and the Not-So-Basics of Technical Analysis

I. Rethinking Trends, Fractals, and a Multi-Timeframe Perspective

Subscribe to 跨市场资产策略笔记

Related content

Related content