There are similarities between the current AI bonanza and the dot-com bubble of the late 1990s but it's not quite the same conditions for a crash yet, says Goldman Sachs.

The conclusion is clear: valuations are high, but AI stocks are not in bubble territory yet.

That's the latest from Goldman Sachs strategists who add another layer to that view: The AI investment boom may have plenty of room to keep going, but echoes of the dot-com bubble are identifiable, so investors should be cognizant of the risks and protect themselves, while still benefiting from the potential upside.

The most frequent analogy analysts draw with the current bonanza in all things AI-related is the dot-com boom and bust of the late 1990s, early 2000s. At the time, leverage rose, balance sheet strength declined, credit spreads widened and equity volatility spiked.

This time around, "the macro and market imbalances... particularly from 1998 onwards... are not generally visible yet," Goldman Sachs asserts in a Sunday report.

Strategists Dominic Wilson and Vickie Chang recommend investors not leave portfolios "excessively vulnerable" to risks of a crash. One suggestion is through options to capitalize on further strength in the sector, perhaps buying call options. Those give an investor the right, but not the obligation to buy an asset at a specified price within a certain time frame. This way investors are only exposed to losing the premium paid for the option and losses are limited.

This boom, while not without precedent, is not on the same scale as previous ones, says Goldman.

They also suggest investors position for wider credit spreads (the incremental return fixed-income investors demand for corporate bonds versus U.S. Treasurys) and higher longer-dated equity volatility, even if the AI boom remains on track for the next year or two.

The strategists declare the "defining feature" of a financial bubble is when asset prices are detached from any notion of fundamental value. They argue that while valuations are elevated, they are not in the danger zone: "Robust earnings have underpinned equity performance."

The 1997-1998 frenzy was characterized by several factors: an investment spending splurge; profitability peaking before the stock rally did; household savings falling as corporate borrowing and leverage rose sharply; crises elsewhere (notably Asian contagion, Russia default, the LTCM collapse); a widening of credit spreads and a jump in stock volatility even as those stocks continued to rise.

Wilson and Chang draw parallels with other bubbles that are relevant like the Japanese real-estate bubble of the late 1980s, and the U.S housing bubble that resulted in the global financial crisis of 2008-09.

They observe that in terms of rising corporate debt (both Meta Platforms and Alphabet have tapped the fixed-income market recently), corporates using up their financial surplus, complex vendor financing schemes (like those with OpenAI, Amazon (AMZN) and Nvidia (NVDA) of late) and the Fed cutting when the economy is not in recession, there are some echoes of 1998.

Profitability, isn't deteriorating yet, they stress. Leverage this time around is more subdued, corporates are still - just - holding on to positive cash balances and household savings are stable. Nor have credit spreads widened significantly, while equity volatility has not reset higher yet.

Thus, they say "recent trends have plenty of scope to continue" and there's "substantial room" still ahead for that AI story.