Silver’s reversal, stocks’ target, and AI

In my yesterday’s free analysis, I wrote that silver's strength was a rally into its vertex reversal, that it was likely temporary, and that the top might already be in.

Silver vertex reversal hits exactly on cue

Today, the market answered. Silver is the weakest part of the sector, down about 1.5%.

There are voices saying that silver could soar to $100 in July, but I wouldn’t bet the farm on that. I mean, sure, silver has 100 reasons to move higher in the long run, but most of them are connected with economic development, and if the stock market finally declines, it’s likely to take silver – temporarily – with it. The upcoming revival is likely to be massive, but if stocks do indeed slide, it might be very difficult for silver to move significantly higher in this environment.

The 1.382 Fibonacci extension based on the 2022 and 2025 price extremes is just ahead (a few points above 7,650). This level could be reached any day – or hour – now.

The excessive bullishness around the stock market and all-things-AI despite crude oil trading close to $100 is breathtaking.

Let me show you two things. One is to provide context and perspective.

Sharp rallies in crude oil that took it above $100 resulted in the biggest declines in stocks. The declines that followed in mining stocks (lowest part of the chart) and silver were huge as well.

I’m not talking about moves that lasted days or weeks. The above chart is based on monthly candlesticks – those moves lower were huge.

This is our context.

The AI hangover is coming

And on the other hand, we have this Nvidia, whose market cap is now bigger than the value of all of entire Australia’s farmland combined. This is just one of the excessive metrics that we have right now.

That’s what bubbles look like. The AI technology is revolutionary. But so was the Internet in general – and yet, the dot-com bubble happened. What’s happening now is a massive AI craze that’s creating a lot of value… for a small minority. 95% of implementations fail to deliver results. Companies make surface-level changes and think they will easily cut costs – and it bites back with vengeance as they learn the hard way how misleading and costly the AI sycophancy can be. It takes time to discover, but the problems are real. Those cool texts, product demos, and everything else that looked great at the first sight, fail to deliver long-term value.

In my view, the true AI strength is in redesigning systems and using it to co-create solutions in a much faster way. However, this requires MORE metal capacity, not less, because what AI produces is seductive – it’s easy to just accept it and call it a day. I know this first hand, because I designed the Gold / Silver StockPicker in this way (the link will provide free access until the end of the week). The thing is that creating it was just 5% of the work. Half of the rest was first making sure that the logic behind what’s being created is correct and is not a result of data mining bias or any other bias. I know how to do this – I was even examined for detecting cognitive and emotional biases during the CFA exams – but can most people say that they even know what those biases are? No. And I’m not writing this to brag – I’m writing this to say that I realize how difficult it is for most people to use AI in a productive way and not in a way that just seems productive.

And the second part of the real workload? It’s being extremely picky with results when the deliverable looks ok at the first sight. How many people do that vs. how many people will just call it a day and then deliver the first iteration of something (application, report, analysis etc.) ready? Some will – the business owners, those with strong ethics, and those that are focused on their mission. Sadly, most people are not among those categories, and they will just deliver the good-looking, but incorrect result.

The valuations seem to take into account the most favorable outcome seen through pink glasses – not the realistic one.

The hangover will likely be painful for many.