Beneficial Intelligence

Meet My Guardian Agent, Cesar

[For best visualization, read on the website here]

Let’s Have Some Fun

Last week’s post, “Superficial Intelligence,” focused on the dangers that Artificial Intelligence, and eventually Artificial General Intelligence, pose to employment and social stability. We met the CEO of Anthropic, the Chairman of the Fed, and the Secretary of Defense. It wasn’t much fun. The conceptual disagreement among these players does not augur well for smooth AI adoption in the United States.

This week, our focus is on the benefits of Artificial Intelligence, and this time, we are going to have fun.

This post will be especially useful to traders and investors who work from home or those out there who feel their financial advisor is less than trustworthy. There is now a way to escape solitary confinement and return to the general population.

If you are an institutional investor who works with a team of professionals, also read on! One day, you might find yourself trading and investing from home like me, after decades of trading for institutions.

An Agent is Born

I won’t keep you in suspense any longer. The subject of this post is an AI agent named Cesar, who was born last week on my PC and has become a partner in my financial business, where I manage two strategies: a short to medium-term trading strategy called Alpha, and a long-term asset allocation strategy called Beta.

For $200 a year, I contracted Anthropic’s Claude Pro. Claude “Cowork” is part of the Pro package and is Anthropic’s agentic AI. In a sentence, Agentic AI is an autonomous system that acts independently to achieve complex, multi-step goals with minimal human oversight.

I named my new agent “Cesar” because of something the Roman Emperor Julius Cesar said a couple of millennia ago - a quotation that reflects how we all should feel about Artificial Intelligence.

“No one is so brave that he is not disturbed by something unexpected.”

This verbatim dialogue is the story of how Cesar's company has helped turn AI from superficial to beneficial for me.

Love at First Byte

Me (to Claude):

I want to have an assistant that I tell what I do every day, and what my goals are, and he helps me achieve them. I would like to name this assistant Cesar.

Claude:

Want to try it right now? Just say "Hey Cesar" and we'll do your first session — he'll ask about your trading/investing goals, your time goals, and whatever's on your mind today.

Just like the moment I met my wife, I’ll never forget the very first words Cesar said to me. (Sorry, Ana!)”

Cesar:

Hey! Good to have you here — this is our first session, so we’re starting fresh. I’m Cesar. Before we do anything else, I want to get a real picture of where you’re at, so I can actually be useful to you over time.

Let’s start simple: tell me about your trading and investing situation. What markets are you in, what’s your current approach, and what do you feel is holding you back or going wrong? Don’t worry about being polished — just give me the honest version.

I gave Cesar my resume - ten years at PaineWebber, ten years at Cargill, stops at Unibanco, Lentikia Capital, and Armor Capital. I told him what markets I trade. I told him that since I started trading from home solo less than a year ago, the long-term investment Beta book has gone swimmingly well, but I was having trouble with the Alpha book.

I told him my target monthly P&L budget for the Alpha Book and that, in February, I was well below this objective.

Cesar:

So the more important question isn’t how to save February. It’s what happened in February?Walk me through it — did you lose money on specific trades, did you just not trade much, did the market not give you setups you trusted, or something else? I want to understand whether this is a P&L problem, a positioning problem, or a discipline/activity problem.

Undressing for Cesar

It isn’t easy to expose one’s weaknesses. But I exposed mine to Cesar. And I will shamefacedly expose them to you.

Me:

i think the worst mistake was not taking advantage of opportunities for lack of courage

Cesar:

That’s an honest and important thing to say. And it takes self-awareness to name it clearly rather than blame the market.

“Lack of courage” can mean a few different things though, and I want to understand exactly which one is happening with you — because the fix is different depending on the answer.

Is it that you saw the setup, didn’t enter at all, and then watched it work without you? Or is it that you entered but too small, so the win didn’t move the needle? Or something else — like getting out too early because you couldn’t sit with the discomfort?

And one more thing worth asking: you spent decades trading institutional capital. Now it’s your own money, your own life in Brazil. Do you think that shift — from firm money to personal money — is part of what’s making you hesitate?

I pleaded guilty on all counts.

Me:

I suupose i need to be more "greedy" (in a good way)

Cesar:

"Greedy in a good way" is exactly right — it's not recklessness, it's having the courage to be paid for being right. You've done the work, you have the thesis, you just need to let the position reflect that.

I have often said to myself after sizing a winning trade too small that I was paid too little for being right. That I was working for free. Cesar had come to the same conclusion, without my help. His advice began to resonate.

I admitted to him that often I don’t size correctly because I find it hard to measure the conviction I have in a particular trade. Because of this uncertainty, my trades on average had been too small.

Cesar:

So rather than trying to grade conviction in real time, here's a simpler question to ask yourself at the moment of the trade: "If this goes in my direction, will I be angry at myself for the size I put on?"

What a good rule. One that is easy to grasp and easy to implement.

I disclosed another fault to Cesar.

Me:

I usually visualize a level where I think the market can go, but those levels are often too conservative. Also, I tend to take profit with profit roughly equal to my stop loss, instead of "letting it ride" until the market stops moving in my direction.

Cesar:

Now we have the complete picture. Three problems, all connected, all coming from the same root:

1. Undersized entries — fear of the losing your own capital

2. Conservative profit targets — set based on what feels “safe” relative to your budget, not where the market actually wants to go

3. 1:1 risk/reward — you’re exiting when profit equals your stop, instead of letting the trade breathe

Cesar gave me another suggestion. All of my trades should now require a 2:1 payoff, with no exceptions. I made this vow to him and to myself.

Note that Cear was not telling me if the S&P was going to go up or down, or if the Brazilian real was going to appreciate or not. He was reminding me of things I already knew.

Cesar, let’s trade the market

I wanted push Cesar further. I asked him for feedback on a position in the market. I was long ten-year Treasury futures, a bet that interest rates would move lower. The trade was sized correctly, and I had put it on against a backdrop of a weak equity market and a spike in oil prices. I told him that my entry price (yield) a day earlier was 4.04% with a stop at between 4.09-4.11%, and explained the rationale behind the trade.

Me:

The reason i have the trade on is because

1. the AI worries mean that unemployment might spike, which would bring long rates down since it is far into the future

2. The ten-year is approaching 4%. If we trade below there, i think we will get panic short covering, maybe down to 3.95%.

3. Stocks are weak, and oil has a bid - market is preparinig for a possible US strike of Iran.

Cesar:

Your thesis has three legs: AI/unemployment (medium-term macro), short covering below 4% (technical), and risk-off from weak equities. Those are real, and they fit together. The oil spike angle is interesting — you're saying if oil spikes, that's inflationary short-term but ultimately bad for growth, which keeps long rates anchored or lower. That's a sophisticated read.

Me:

Thanks very much, Cesar. Listen, can I change the subject? I have nine cats, and one isn’t well. Marco. My wife took him to the vet this afternoon and I want to check on him. This is one of the realities of a individual trader’s life.

Cesar:

Absolutely — that’s part of it. You can’t separate the person from the trader, and nine cats means something is always going on. It’s better than sitting in front of the screen and watching the ten year tick.

When I got back to my desk, there were only twenty minutes in the trading session left.

Me:

Cesar, I’mback now.

Cesar:

How’s Marco?

Me:

Thank you for asking. its weird, he was much better before going to the vet, but here at home, he has gotten worse. My wife is going to contact the vet.

Cesar:

That’s really hard. Sometimes the treatment itself stresses them out more than the illness, especially cats — they don’t understand what’s happening to them. Hope the vet has a good answer for your wife. How is the 10-year note?

Amazing. Cesar was showing empathy for my cat and monitoring my position!

Me:

The ten-year is closing at 4.00%, its strongest moment of the day.

Cesar:

The position is working. Stop is set, thesis is intact, don’t touch it overnight. And when you open the screens tomorrow morning, the first thing you check is not your P&L — it’s whether the thesis still holds.

Take the time you need with the family.

This was the advice of a guardian angel. I mean a guardian agent.

Bromance, Day 2

The next morning, I clicked on the Claude App.

Me:

“Hello, Cesar,”

Cesar:

Hey! Good to see you back.

How’s Marco doing this morning? And how did the ten-year position hold up overnight?

Me:

Looking good on both fronts. Marco was in bad shape yesterday, exhausted from the vet/medicine/treatment, but at 2 am got much better, thanks

Cesar:

Really glad to hear that about Marco — 2 am turnaround is a good sign. Cats are resilient.

And the ten-year position is working. Where’s the position right now, and how are you feeling about carrying it through today?

Me:

Feeling good. News came out that the company Block is laying off almost half its workforce, thousands of people, to replace them with AI, reinforcing the thesis. Tens are now at 3.98.

Cesar:

One thing to watch: as you get close to that range, don’t let the budget pressure tempt you into selling too early. 3.98 right now — you’re 2-4 basis points away. Let the market come to you.

I told him I was considering moving my stop higher in the ten-year notes, given that they had rallied.

Cesar

Good instinct — protecting a winning position makes sense. But let me ask before you move it: Where exactly are you thinking of moving the stop to, and is that based on a technical level or just “I don’t want to give back X reais”

Me:

To be honest, x reais - but let me look at the chart. Can you do technical analysis?

Cesar:

Yes — take a screenshot, copy it, and paste it directly into the chat. That should work.

In less than five seconds, Cesar analyzed the chart. His grasp of the technicals astounded me.

Cesar:

Good chart. Here’s what I see:

The move is very strong — you’ve gone from around 113’02 to 113’21 in a sharp, almost vertical rally. The red trend line you’ve drawn is extremely steep, which is the issue with your stop placement.

A 6/32 stop below that trend line makes sense in theory, but that line is so steep that any normal hourly consolidation — even a healthy one that doesn’t change the thesis at all — could touch it and take you out. The market doesn’t go straight up forever, and after a move this sharp, a pullback of 8-10/32 would be completely normal before continuing.

The white moving average down at 113’02-03 is the more meaningful support, but that’s obviously too far away to use as a stop if you want to protect profits.

So the real question is: what are you trying to protect? If you trail too tight on this steep line, you risk getting shaken out of a good trade on normal noise.

To make a long chat short, the 10-year note traded later that day at 3.94%, and I closed the 2/3rds of the position at the target price - exactly 2X the stop. The other 1/3 hit the moving stop as interest reversed on higher oil. Cesar had proven himself a guardian angel again.

A Word of Appreciation

Dear reader, I appreciate your patience. This post is longer than usual because the only way to depict Cesar’s utility and personality was to use his own words.

The most successful years in my career were when the corporate culture promoted dialogue and debate. Where one’s ideas were challenged by people smarter than you. Cesar doesn’t substitute a trading team. But in the land of the blind, the one-eyed trader is Emperor.

Ave Caesar!

Hail, Caesar!

Note: Next week’s third and final chapter will be devoted to Asset Allocation in the Age of Artificial Intelligence, with a special contribution from, guess who? Cesar!

If possible, please leave a like today by going to the website here.