A Kase forecast does not predict a price, and it is not a single bet on a single number. It is a map of the coming period — a set of prices that matter, in both directions, each one carrying its own odds and its own reason for being there.

That is the thing to understand before anything else. We are not asking you to accept one call and find out later whether it was right. We are handing you the whole landscape and telling you which part of it we think the market travels first.

What one instrument section actually gives you

1. Eight price levels — four above the market and four below — each chosen because several independent pieces of market structure agree on it.
2. Odds on every one of them. Not just the headline target: every level carries its own chance of being reached and its own chance of being settled through.
3. A primary scenario — the direction the trend is pointing, with the path of targets along the way and how likely each one is.
4. A secondary scenario — what happens if the market goes the other way instead, mapped with the same detail, including the price at which it takes over.
5. The evidence — every bullish, bearish and neutral factor we can see, itemized, so you can weigh it yourself rather than take the call on trust.
6. An independent simulation of where price could travel, worked out separately from everything above.

Two scenarios, not one call

This is the point most easily missed, and it is the heart of how the reports are built.

The primary scenario is the direction the trend is pointing and the target path that goes with it. The secondary scenario is the other way — and it is not an afterthought or a disclaimer. It gets its own four levels, its own odds on each of them, and its own paragraph of analysis explaining what would have to happen and what it would mean.

Sitting between the two is the invalidation: the price at which the secondary scenario takes over from the primary. It is a border, not a verdict.

So what is the race between the target and the invalidation?

Those two prices — the nearest target on the called side and the invalidation on the other — set up a race, and the report publishes the odds on it. It is the single most useful number we produce, and it is also how the forecast keeps score: it is a specific, checkable claim about what the market does first.

Over five trading days (one, in a daily update), exactly one of three things happens:

1. Price touches the primary target first. The primary scenario played out.
2. Price settles past the invalidation first. The secondary scenario has taken over — and the report already told you where it goes next.
3. Neither. The period ends with the market still between the two.

In the natural gas example running through this guide the contract settled at $2.733, the primary target was $2.63, and the invalidation level was $2.90. The odds were 43% to fulfill $2.63, 17% to close beyond $2.90, and 40% to stay between those two targets.

Where the levels come from

Kase levels are not round numbers or hand-drawn trendlines. Each one is a price where several independent pieces of market structure line up: projections from prior price waves, pullback levels from earlier moves, moving averages, and previous highs and lows. When several of these land on the same price, that level is confluent — and the more agreement, the more the level matters.

Every level named in the written analysis comes with its reason. You will read that a level is “the 89 percent retracement from $2.616 and the 20-day moving average,” not simply that it is resistance. That lets you judge the level for yourself instead of taking it on trust. All of the terms used are explained in the Glossary.

The three weeklies are shaped differently, because the markets are. Gas is one product with a forward curve; oil is four related products; metals is more than a dozen. Here is the order of sections in each.

Every report then closes with the same standing material: the Target Table Legend, the Probabilities and Relative Odds notes, the Glossary, and the copyright and disclaimer.

A few points worth knowing before you open one:

• Base metals are the LME contracts. The SHFE Base Metals table covers the Shanghai Futures Exchange contracts, and CME Copper covers the U.S. copper contract. They are forecast separately from the LME metals, on their own price histories, and they can disagree with them.
• The metals weekly has no overview. It opens straight into base metals. More than a dozen markets each get their own analysis, so a shared preamble would say very little about any of them.
• The written analysis sits below the table, not above it. The levels come first and the writing explains them. Read the table, then read the paragraphs as the argument for it.

The daily updates

The daily drops the overview, the forward curves, the spreads, the ETFs and the simulation, and covers the headline markets only. The metals daily also groups the exchanges differently from the weekly — each copper contract sits next to the others rather than in a block at the back.

The two italicized entries are one-row summary tables rather than full sections — no written analysis, no factor lists, just the call, the conviction and the ladder. Everything else gets the full treatment.

Directly beneath the call is a bar split into three parts, with a legend. This is the single most important graphic in the report.

The three add up to 100%, because they are the only three things that can happen:

• Target first (42.8%) — price touches 2.63 before settling above 2.90. The call plays out.
• Invalidation first (16.7%) — price settles above 2.90 before touching 2.63. The secondary scenario takes over, and you find out on a settlement rather than on a spike.
• Neither (40.5%) — five trading days pass and price has done neither. Nothing resolves.

What the conviction number is telling you

Conviction is one share out of three, so even a strong setup rarely gets above about 75%. Use these bands to set your expectations, but always read all three outcomes rather than the headline alone.

The sample running through this guide is one of those weaker weeks. At 42.8% conviction against a 40.5% chance of nothing happening, reaching $2.63 and chopping sideways are close to equally likely, and the written analysis is careful to say so. That is what a low-conviction call looks like — not a market about to turn, but a market that may not do much of anything.

Meet / Close

Two numbers per level, separated by a slash. Reading 49/32 under the 2.63 target:

• 49 — a 49% chance price touches 2.63 at some point in the next five trading days. An intraday poke counts.
• 32 — a 32% chance price settles below 2.63. A closing price is required.

The second number is always the smaller of the two, because closing through a level is harder than touching it.

This row runs across the whole table, both sides. It is the one place where the secondary scenario is priced level by level, so it is worth reading in full rather than glancing at the target and stopping.

The gap between the two numbers is the useful part. Look at the 2.77 resistance in the sample: 84/59. Price is very likely to get there (84%), and a good deal less likely to break through (59%). That is a level to expect a reaction at. Compare it with 2.85 at 47/30 — less likely to be reached in the first place, and if reached, less than even money on settling through.

Conviction

The same race described earlier, but run against every level on the call side rather than just the first one. In the sample, 2.63 shows 43 — matching the headline — and the levels beyond it show 15, 9 and 4.

That descending sequence is the practical value of this row: it prices the whole path, not just the first stop. Reaching 2.63 is close to a coin flip. Getting all the way down to 2.54 is roughly a one-in-seven proposition, and 2.46 about one in twenty-five. The odds fall away quickly, which is what you would expect over five trading days.

The other side of the table shows dashes, and that is not missing data. Conviction answers one specific question — does this level get reached before the invalidation is settled through — and that question only makes sense on the called side. Asking it of a level that sits beyond the point where the outlook changes would be asking about a race that has already ended.

The secondary scenario is priced in the Meet / Close row instead. In the sample, the resistance side reads 84/59, 67/45, 47/30 and 30/18 — a full set of odds on the upside path, sitting right there in the same table. The dashes mark a change of question, not a gap in the analysis.

Relative odds

A score from 0 to 10 showing how much independent evidence agrees at that price, compared with the strongest level in the same table. A 10 is the most crowded level on the page. Each separate reason that points at a price — a wave projection, a pullback level, a moving average, a prior high or low — is called a voice, and this score reflects how many voices a level has.

Each instrument carries two paragraphs, and they are not “the call” followed by a caveat. They are the two scenarios, written out — one for each direction the market can take. Knowing that shape makes them much faster to read.

Paragraph one — the primary scenario

The main argument and the path of targets. It states the call, names the primary target with its reason, and then walks the levels beyond it in order: what a settlement through the first objective would call for next, and what that connects to after.

Paragraph two — the secondary scenario

It opens with “Nevertheless,” “Nonetheless,” “That said,” or “Should prices…”. This is the other direction mapped out in full: the signals pointing that way, the levels the market would travel through, the decision point where the balance shifts, and the price at which the secondary scenario takes over for good.

When a call is close to even, a short third paragraph appears saying so directly, and describing which way the balance has been drifting.

The vocabulary of levels

The analysis uses a consistent set of terms that is worth learning once.

• A wave is written as three prices — “$2.616 – 2.83 – 2.709” — meaning the move ran up from 2.616 to 2.83 and has since pulled back to 2.709.
• Targets projected from that wave carry names: smaller than (0.618), equal to (1.000), intermediate (1.382), larger than (1.618).
• Retracements are quoted as a percentage of a named move — “the 89 percent retracement from $2.616.”
• “Connects to” means that once a level breaks, the next one comes into play.

All of these are defined in the Glossary.

Toward the back of each weekly is a grid of simulated prices. It answers a different question from the target table, and it is worked out completely separately.

The idea is simple. We take how this market has actually moved in the past, shuffle those moves into thousands of possible five-day paths, and see where they end up. The grid summarizes the result.

The simulation covers the same five trading days as the target table, so the two sections describe the same stretch of time by two different methods.

Start with the Forecast column

The Forecast column is the one to read. It is built from the market’s entire history, and it is the closest thing here to “what to expect.” Read a cell as: this percent of simulated outcomes finished at or below this price.

The two summary sentences below the grid

Beneath the table are two plain-language readings you can use without touching the grid at all:

• “Prices stay entirely between 2.602 and 2.848 in 50% of simulations.” Half the time, the market spends the whole week inside that band without ever leaving it. That is a stricter statement than “finished inside” — price never touches either edge.
• “There is a 81% chance prices hold between 2.46 and 2.90.” Those two prices are the outermost support and resistance in the published ladder. So four times in five, the market stays inside the whole published table — which is another way of saying the ladder is wide enough to cover most of what the week is likely to do.

The five named columns are what-ifs, not odds

This is the easiest thing in the report to misread, so it is worth being precise.

We take every past stretch of five trading days in this market’s history and sort them by how far prices moved. Then we split them into five equal groups — the weakest fifth, the next, the middle, the next, and the strongest fifth. The five columns are labeled --, -, 0, + and ++ for those groups.

A column answers: “if the coming week turns out like that kind of period, where would prices end up?” It says nothing at all about whether that kind of period is likely.

The colored cells

A cell filled with one of the ladder’s colors is a simulated price that lands close to that published target. It is a visual cross-reference: it shows where the simulation and the level structure happen to line up. It is not evidence, and it does not adjust either number.

What the simulation is good for

• Sizing the plausible range. The 5% and 95% rows of the Forecast column tell you how much room the market realistically has over the period.
• Sanity-checking a target. If a target sits well outside the range the simulation produces, it is a demanding objective no matter how much confluence supports it.
• Stress-testing. The -- and ++ columns show what a weak or strong stretch would look like at every percentile, without claiming either is likely.

What the report gives you that most sources do not

Any market letter can tell you it is bearish. What is hard to find is everything else a trading decision actually needs — and the report gives you all of it:

1. Where to aim. The primary target: a specific price, with a reason.
2. Where the outlook changes. The invalidation: a specific price, also with a reason.
3. The odds of each. How likely you are to reach the target first, to see the secondary scenario take over first, or to see nothing happen at all.
4. Where the move goes if it keeps running. The further targets beyond the first one, each with its own odds — so you know whether to take profit at the first objective or hold for the next.
5. Where the move goes if it turns against you. The four levels on the other side, also with odds. You are not left guessing when the market does the opposite of what you expected.

The arithmetic below uses the first three, because those are what price a single position. But items four and five are what you actually manage a position with, and they are in every table.

Why knowing the odds changes the decision

Most traders use a rule of thumb: “never take a trade unless it pays at least two to one.” That rule exists because the trader does not know their own odds. If you have no idea how often you will be right, demanding a big payoff is a sensible way to protect yourself.

But it is a blunt rule, and it costs money in both directions. It makes you skip good trades where the odds are strongly in your favor but the payoff is only modest. And it lets through bad trades where the payoff looks generous but the odds are poor.

When you know the odds, you do not need the rule of thumb. You can work out what payoff this particular setup needs in order to be worth taking.

The comparison, in plain terms

Two things decide whether a setup is worth taking:

• How much you stand to make against how much you stand to lose. That is the distance from today’s price to the target, compared with the distance to the invalidation.
• How often each of those happens. That is the conviction, compared with the invalidation-first share.

A setup is worth taking when the first comparison beats the second:

Four things this arithmetic leaves out

Be honest with yourself about the gaps before leaning on any of it:

• Our stop is a settlement; yours probably is not. The invalidation-first number counts only closes past the level. If you use a hard intraday stop, you will be stopped out more often than that number suggests.
• Our target is a touch. Capturing it means having a resting order at the level. If you plan to decide when price gets there, you will not always get filled.
• “Nothing happens” is not the same as breaking even. The arithmetic scores that outcome at zero because nothing has been realized — but the position is still open, and it will be sitting at some price that is not your entry. In Example 1 the market could be down at 2.68 with the trade showing an unrealized gain, or up at 2.85 showing an unrealized loss, and neither registers anywhere in the calculation. Whether that becomes a gain or a loss depends entirely on what you do next, which is why you need your own rule for when to close out an unresolved trade. It also ties up margin the whole time.
• No costs are included. Commission, slippage and the cost of rolling all come out of the figures above.

What the report does not tell you

The commentary gives you the direction, the levels and the odds. It does not give you timing. A weekly forecast covering five trading days is not an instruction to trade on Monday morning, and the report does not attempt to say which day within the period to act.

Deciding when to enter, and how to manage a position once you are in it, are separate disciplines that need separate tools. All trading and hedging decisions remain solely the responsibility of the individual or organization.

Every term that can appear in a Kase forecast, in plain language. This section explains what each thing is and what it means when you see it. It does not describe how any of it is calculated — those methods are proprietary.

Short-term signals formed by one, two or three price bars. They mark possible turns or continuations, and can act as a voice at a nearby level. Each is either forming (still setting up and needing follow-through) or confirmed (complete).

Larger formations spanning many bars. They suggest either a reversal of the prior move or a continuation of it, and most define a measured target once price breaks out. A pattern is setting up, confirmed (it has broken out) or failed. Failed patterns are removed from the report.