Milton Berg: I Have Evidence Market Has Likely Bottomed | Why Milton’s Long Semis, Korea, Nasdaq, and More (With Caveats), and Why He Thinks Gold has made a Multi-year Top

Milton Berg’s evidence points to a market bottom, but not yet to an assured new bull market. He cited at least 34 buy signals in late March and April and went long after the July 29 low, following a period short. The S&P 500 held above its June low even as the Nasdaq 100 and Philadelphia Semiconductor Index made new lows, a divergence he viewed as supportive. A July 29 cycle date and heavy retail selling added to the case for a low. The rally from the March 30 lows had already lifted the Philadelphia Semiconductor Index more than 100% and the S&P Technology Index more than 90%.

Berg treats his signals as evidence for being invested, not as promises of a particular return. Several April patterns had limited prior histories, and he discounted the large projections from a signal with only one precedent. For other signals, he cited historical minimum and median gains, including a combined April-signal minimum of 6.82% for the S&P 500 and a median maximum of 15.49%; those are historical outcomes, not guaranteed targets. Most signals were behaving in line with precedent, but an April 27 Nasdaq signal was not. A buy signal remains valid if the original low holds, he said, even if the market pulls back to test it.

After covering his short and buying on July 29 and 30, Berg’s portfolios were fully invested without leverage. His allocation was nearly 50% SPY, 20% the Philadelphia Semiconductor Index, 10% the Nasdaq 100, nearly 10% the Russell 2000, and 5% each in S&P mid-caps and Korea’s KOSPI. He saw the sharp selloffs in semiconductors and Korean shares as potential panic selling and expected at least a short-term bounce. But he considered a retest of their lows within a month or two quite possible; he viewed a multi-month Korean rally as the most logical outcome only after allowing for that risk.

The KOSPI had fallen 43.93% in 27 days into its July 29 low, while the semiconductor index had dropped 28.73% from its June peak. For Korea, Berg said a retest could make a new closing low without making a new intraday low, as in his 1987 comparison. A break below the July 29 intraday low would be very bearish; from the August 10 level, that low was 19% lower, versus 12% to the closing low. A decline within that range was possible in his view, while a drop beyond 19% could indicate much deeper downside. He also saw exhaustion risks in the Nasdaq 100 and semiconductors, including gaps without upside follow-through, and was watching for a short-term top.

Berg said broad participation did not confirm the strength in the S&P 500 and Russell 2000, and he acknowledged that a sharp rally could be only one leg up. A retest of crash lows would not be unusual: he noted that after the 1987 decline the market later came within about 1% of its low, though the COVID decline instead formed a V-shaped bottom. He had no evidence that the current decline had become a bear market, but said a break below the lows was an outside chance. Extreme valuations, high margin debt relative to cash, and leverage in the economy concerned him, yet he was not forecasting a major bear market; he noted that major bear markets usually follow Federal Reserve tightening, which had not occurred.

Berg is negative on gold and silver over the long term, believing gold likely made a multi-year top in January, but he distinguished that view from a call on a near-term trading top. He said he sold gold on January 29 after concluding the top had arrived, though not at the exact high, and later remained out of gold, silver, and GDX after exiting before a rally. He sees possible further rallies but does not expect new highs, citing gold’s exceptionally high value relative to crude, housing, and other commodities. He said the 1980 peak in gold relative to inflation preceded a 20-year bear market, while cautioning that valuation extremes alone are not trading signals.