vorin

IB front-office dev turned quant-trader, unleashing algos on my personal portfolio while being constantly reminded that markets are harder to debug than code

[kitze the 🐐 @thekitze]

anthropic including fable in the max plan

[Psyho @FakePsyho]

Quick summary of what happened during AWTF in Japan:
- OpenAI crushed humans in both Heuristic and Algorithm categories. Heuristic is the one where I won in 2025 and OpenAI placed 2nd
- AWTF is pretty much the highest level competitive programming contest: invited finalists in both categories are among the best people in the world; this is essentially the first time AI won vs humans in a programming competition in such a decisive matter
- big change compared to 2025 is that AI was constantly progressing and didn't get stuck / plateau at any point
- System used for Algorithm is mostly just custom harness + model very close to gpt 5.6; my educated guess is that system used for Heuristic is a custom "autoresearch" harness with a swarm of cooperating agents and a massive inference cost
- OpenAI was sure that they will win Algorithm and were very confident that they will win Heuristic; they did a lot of backtesting on old contests and they would win all of them in their simulations
- imho heuristic problems are a great proxy for ML autoresearch capabilities; if AI was able to match best humans here, we're very close to RSI / automated researcher; this result is way bigger than a high score on some questionable benchmark

When I get some rest, I'll try to post more thoughts about the whole thing.

Thanks for following the coverage. I believe I tweeted more in those few days than I did in the first half of 2026. And I'm not even counting the 10 hour long livestream.

Oh and technically speaking, I didn't participate so I'm still undefeated and I'll gladly keep "Humanity's Last Programmer" in my bio.

[vorin @V0rin]

@FakePsyho Was OpenAI using just one agent at a time?

[vorin @V0rin]

@OddStats Ideally both 👌

[rekt @rekt_tekashi]

@michaeljburry all a bear can do
is just vibe to it all

[OpenRouter @OpenRouter]

Introducing the Fusion API, the smartest compound model in the market.

Fusion achieves Fable-level intelligence at half the price.

How it works 👇

[Z @ZeeContrarian1]

The worst thing that can happen to the market right now is actually a peace deal.

A real peace deal gets you one 3% up day.

“Sources say a peace deal is very close” gets you a 3% up day every week.

[VolumeLeaders @VolumeLeaders]

$SPY / $SSO - I get this question almost every day.

"Buy or sell"
"Bullish or bearish."
"Up or down."

“I have a job / family / life / tee time, man. Just tell me what to do.”

The short answer is that there's no 'easy' button, no free lunch, no guarantees. And there's certainly nothing in the data that says "hey man, Citadel backed up the truck here and you should too." So let's stop seeking shortcuts. There aren't any.

The longer answer is below. Get comfortable if you decide to keep reading. I wrote for three hours.

Ok.

I’m probably not going to tweet “buy” or “sell” because I just don’t know. Not for a fact.
I can give you my opinion when I have time to form one, and I sometimes do, but it’s still just an opinion. And as the adage goes, opinions are like nipples. We all have them, they’re sometimes fun to look at, and most of the time they don’t do anything.

The truth is, I just don’t have 88 hours a day to analyze 25,000 trades and provide high confidence reads for which ones I think are bullish and which ones I think are bearish and why. Sometimes I just post a few charts between meetings and hope you guys can run with them.

I’m probably not going to say “bullish” or “bearish” either because there are too many time horizons to consider. Nothing is always bullish on all timeframes. Nothing is always bearish on all timeframes. I don’t know what the heck time horizon you’re trading just like you don’t know mine, so I’m not going to say “this is bullish” and then have some guy come yelling at me in my DMs because he trades on the 1min chart and I am thinking about what’s going to happen in the next few months.

Case and point – the $IGV accumulation from February to April. I probably posted 10 times about those trades. It was bullish on longer timeframes, but there were plenty of times when longs had to eat shit for a few days while price meandered in a range.
So, were those prints bullish? Yes. Did price always do bullish things always no matter what always forever and ever always? NO. And price didn’t even do anything remotely bullish for several weeks, despite the accumulation we were all watching unfold.

So…
If I say “it’s a buy” and you go buy 0dte calls as price declines, you’ll get wrecked.
If I say “these are bullish” and you take a 3% drawdown the next day, you might be pissed because in your head, bullish means up and price didn’t go up.

“Bro, you said it was bullish.”
“Bro, you said up.”
“Bro, it didn’t go up.
“Bro.”

What’s bullish for a guy swinging shares on a multi-week timeframe is entirely different from what’s bullish for a guy buying 0dte calls from the bathroom stall at his job.

So, let's all stop asking "buy" or "sell." You can't know. I can't know. It's always a guess. It can be an educated guess, it can be an informed guess, but it's still a guess. And even if we're right, it doesn't mean price will respond immediately.

That said, sometimes I can find a lot of confluence between a given trade and other factors. Other times I can't. But the purpose of this exercise is not to be spoon-fed a signal that says "buy" or "sell" and then leap into an irresponsible trade. Things just don't work that way. You're going to have to do a little work. You're going to have to look at more than one thing. VL doesn't spit out trade recommendations. It describes institutional positions and provides context for them. It doesn’t tell you what price is going to do on the next 5-minute candle.

Just like when your favorite indicator flashes "buy." It's not a guarantee that price will move the way you want it to, exactly when you want it to, and go as far as you'd like it to go. Trading is a game of probability. Your favorite indicator works sometimes. Other times it doesn't. So, you overlay it with other indicators to get more confluence. If you have five indicators that all say "buy," your odds of success are better than if there's only one.

VL is another layer to that process. That's all.
You take what you have, you add VL to it, and your hit rate goes up. You find more good opportunities, and you chase fewer bad ones.

So - back to the opportunity.

You're at your desk. You have your VL dashboard up. SSO #10 comes in.

Now what.
Now fucking what.
What the hell do I do with this big stupid orange bubble that has a number on it instead of the word "BUY" or the word "SELL"
How do I know what to do?
If nobody is telling me what to do, how will I know what to do? Will Grok tell me? What do I do if Grok won’t tell me!?

YOU HAVE TO DECIDE.
Consider the data you have. Then do your best.

VL is not going to tell you when to take a trade. But it is going to provide you with valuable, contextual information on trades that were already taken, and that information can help guide your decisions. It just won't make them for you.

Among the many things it does is help you find opportunities you may not have known about (such as this one).
But it's not a stand-alone trade generator. It's an information platform that provides facts.
So, let's review those facts.

- You see an unusual position arrive.
- You have context for that position which indicates exactly how large/rare/important it is. What’s context? Context is its numerical trade rank, its percentile ranking, its relative size, its last comp, its frequency.

Those elements describe the trade in ways that raw data cannot. They describe IMPORTANCE. At the end of the day – we all just want to know if this trade is important enough to pay attention to. In this case, the answer was yes. So much so that I spammed multiple rooms in the VL Discord to ensure everybody saw it, something I very rarely do.

You also have nearby levels, trade clusters, trade attributes, and so on. Lots of ‘things’ that VL tracks to help you better understand what just happened. And with tickers like $SSO, you can also look at about two dozen $SPY surrogates – all of which have been quite active these last few days. $SPY, $SPYM, $VOO, $IVV, and others. All dealing with the same underlying instrument – the S&P 500. Combine them. Look for patterns.

- You can infer that this position will be defended. What do I mean by defended? I mean that if I just put $300M trade on the books in a leveraged ETF, I’m going to darn well ensure that it doesn’t lose money. I am going enter at the right time, and I will keep price trading at or above it for as long as possible. That may mean throwing more money down in other (probably smaller) trades to keep price afloat. But nobody puts that kind of money on a trade and then just crosses their fingers and hopes for the best.

Size = Confidence. This trade had BDE…and some.

- You have charts (VL charts and your own) that can show you any number of additional data points such as the one described below. If you like the 50DMA, great. If you have others too, even better. The more confluence you can find among data points, the higher confidence your entries and exits can be. So, look the trade in VL, then look at your charts, then look at whatever else you look at, maybe bounce the idea off other folks in the community, and then let all that information swirl around in your brain for a while until you can arrive at a decision. If you have enough confluence to take a trade (or close an existing position), do it. If not, move on. VL tracks thousands of trades every day across 5700 tickers. There are plenty of fish. I mean that.

How did I interpret it? What was my thought process? Well, I use ATR bands among other things. It's not the only way to do things, it's just one method among many. Spend some time in the VL Discord and you'll see all manner of traders with any number of different strategies. They're all unique, but the one common thread is that VL improves all of them because instead of relying solely on price, volume, fibs, lines, channels, and 50 different oscillators, VL tells you when, where, and how real, actual money is being put to use by the largest and most influential players in the market.

So, here's an hourly chart for $SSO. The white arrow is where price was when $SSO #10 arrived. Price had been declining for several sessions. It was roughly two ATR below the 200sma (yellow line).

If you don't know ATRs, that’s fine. They're like standard deviations. Not exactly, but close enough. Price gets too far from the mean and it reverts to it. Then it gets too far away in the opposite direction, and it reverts again. The point is that price was in an area (green ribbons) in which you'd expect to see buying, not selling.
Buy low, sell high.
If price is extended below the mean and long leverage arrives, odds favor buying over selling.

It's not a lock.
You can't bet the farm on it.

You still have to turn on your brain, think about all the variables at play, size appropriately, and manage the trade if you decide to take one. But by overlaying the VL data with a “regular” chart, you can increase the confidence of your read. Even if you were positively clueless about what to expect, you can see a trade like this arrive and know enough to close or protect short positions if you had them. Size like this comes with magnitude and the last thing you want when short is to get your doors blown off by a ‘surprise’ – which is exactly what we got.

Anyway, I had been thinking about this same hourly chart in the last 24-48 hours ($ES instead of $SSO). I discovered a channel (sort of by accident) that suggested price was respecting the upper and lower bounds of those two green lower bands on the hourly timeframe. So I was thinking a move back to the upper band around 7400 would be appropriate given that price was near the bottom of the lower band. I didn’t know that big $SSO trade was going to come. I didn’t know we’d get a rally today. I didn’t know squadoosh. I just saw a pattern.

When overlayed with the $SSO positioning though, I had more clarity. Not absolute clarity. Not guaranteed success. Just…better odds of interpreting it correctly than I had before.

In conclusion, the VL platform is there to help. It’s meant to complement your strategy, not replace it. Bolt it on to whatever you’re already doing, and your odds of success improve.
https://t.co/i5V7d5Evfk

[vorin @V0rin]

@TimJDillon The things that's interesting to me is if you consider how much of a positive outsider and pivotal person Massie is, a few wealthy people could've easily donated $30m to him, but they didn't, what does it say about the reality?

[vorin @V0rin]

@OddStats gold!

[vorin @V0rin]

@BlueHorse88 🤦🏻

[vorin @V0rin]

@BlueHorse88 True, you're right they can't go hardcore until AI is not doing at least 50% of stuff without crashing the economy, my guess until then give the laid off people jobs to build up manufacturing infra needed later for AI and data centers ofc lol, that could work

[vorin @V0rin]

@citrinowicz I'm curious what's your take on the issue of trust. Will Iran accept terms, however generous, that are guaranteed only by the US? I don't imagine Pakistan has enough weight here

[vorin @V0rin]

@profplum99 There might be at least some short-term gains, but do you think there's a line where doing the right thing or not meets practicality in long-term gains. I mean undercutting allies, burning trust, in your face everything for me but nothing for thee and if it's been crossed

[Nic @puckrin]

Wow.

A former US Treasury Secretary suggested US authorities prepare a back-up plan in case of a collapse in demand for Treasuries.

This is the same guy who served during the Great Financial Crisis.

Nothing to see here...

[vorin @V0rin]

@Bogachan_1971 As perverse as it seems I'm leaning towards that they want to continue with the rally for a proper blow off top and they need some wall of worry to climb, before the reality hits the markets properly in a month or 2

[vorin @V0rin]

@TradesByC I think it's a weird way of showing a contract roll from May to June

[Trita Parsi @tparsi]

Why the Iran ceasefire may have shifted the dynamics back in Trump's favor

Diplomacy between Washington and Tehran has not yet unraveled, despite JD Vance’s theatrical departure from last week’s talks in Islamabad. Trump now signals that the two sides could reconvene within days in the Pakistani capital. Whether negotiators return to the table or continue their exchanges through quieter, remote channels before the ceasefire lapses, one reality appears to have shifted: Trump has clawed back a measure of momentum—and with it, leverage—over Iran, largely by virtue of the ceasefire. Here’s why.

Trump entered this moment politically cornered and strategically constrained. Surging gasoline prices were inflicting acute domestic pain, eroding his standing at home. More critically, he faced a barren escalation ladder. Each conceivable move—strikes on Iran’s oil infrastructure, attacks on civilian targets, the seizure of Persian Gulf islands, or covert operations to capture enriched uranium—carried the near-certainty of forceful Iranian retaliation. Such responses would not merely match his escalation but compound it, deepening his economic exposure, amplifying political risk, and entangling him further in a perilous and unwinnable strategic bind.

Nor could he simply extricate the United States from the conflict on his own terms. Absent an understanding with Tehran, Iran retained both the capacity and the incentive to continue targeting Israel and vulnerable U.S. assets across the Gulf. Trump needed Iran’s permission to get out of the war.

The ceasefire, however, has subtly altered that equation. Trump may no longer need a formal nod from Tehran to step back. If he disengages now—without a comprehensive agreement—Iran will almost certainly maintain its grip over the Strait of Hormuz, a strategic setback for Washington. Yet Tehran is unlikely to resume direct military operations against U.S. targets in the Persian Gulf. To do so, in the absence of renewed American strikes, would cast Iran as the aggressor, inviting severe and potentially coordinated repercussions—not only from Washington but from wary global powers such as Russia and China.

Moreover, the balance of needs has tilted. Iran now appears to need an agreement more than the United States does. Trump has already secured his central objective—the escape from a war he was ill-advised to begin—while Iran, despite accruing leverage through its command of the Strait, remains far from realizing its broader ambitions: meaningful sanctions relief, a definitive and enduring end to hostilities, and perhaps even the contours of a more stable, constructive relationship with Washington.

Tehran’s decision to dispatch its largest, most senior, and most expansive delegation to Islamabad for direct talks with the American vice president reflected a striking confidence—that it occupied its strongest negotiating position vis-à-vis the United States since 1979. Yet to convert that moment of perceived ascendancy into little more than a cessation of U.S. bombardment would fall short of its aspirations. Even if Washington were to acquiesce to Iran’s control of the Strait, such an outcome would pale against the far more consequential gains Tehran believes are within reach.

Instead, Iran needs to translate this leverage not only into a durable end to the war, but ideally, into a new peace: One that delivers sweeping sanctions relief and inaugurates a more stable, mutually defined economic and political relationship with Washington. Such an arrangement would serve as a bulwark against renewed conflict. The economic imperative is especially stark: sanctions relief is indispensable to reconstruct a country now burdened with damage running into the hundreds of billions of dollars.

As I have argued before, sanctions relief is not merely an economic demand—it is a strategic necessity. Without it, Iran risks a condition of chronic erosion, a slow but steady weakening that would leave it exposed. That vulnerability, in turn, could invite further attacks. It was, after all, the misperception of Iranian weakness that helped open the window for initial strikes.

But Trump does not, in any fundamental sense, require any of this. The United States can endure without a formal agreement with Iran and without the benefits of an economic relationship with Tehran. To be sure, a negotiated settlement would better serve long-term American interests: the nuclear constraints Trump seeks can only be credibly secured at the negotiating table. Abruptly abandoning diplomacy while leaving Iran in undisputed control of the Strait would also unsettle key regional allies. Yet these are strategic preferences, not immediate necessities.

Trump’s calculus is far more transactional and far less patient. He can point to the damage already inflicted on Iran’s nuclear infrastructure and conventional forces, proclaim a hollow victory, and disengage. He has already emphasized that the United States no longer depends on Persian Gulf oil, insulating it from the direct economic consequences of Iran’s toll regime. As a result, the burden shifts outward: the Strait becomes a problem for European and Asian powers—countries that Trump has noted declined to rally to his side when he sought their help in prying the waterway from Tehran’s grip.

The window now open offers Tehran a chance to convert battlefield leverage into lasting strategic gain. To let it close would mean forfeiting not just incremental progress, but the possibility of reshaping its economic and geopolitical position. By contrast, the United States, having already secured a tenuous exit ramp through the ceasefire, has less at stake in the short term. Walking away, therefore, is politically and strategically easier for Trump than for his Iranian counterparts. Both can live with diplomatic failure, but Tehran has more gains to lose.

How Tehran chooses to navigate this narrowing corridor—whether it presses its advantage or overplays its hand—will be interesting to see.

[Ed O'Malley @EdwardOMalley]

Here is Trump as a pilot.