I argue that this is what has changed:
"Treasuries have also benefitted from the wider adoption of non-cash collateral since the crisis. Just over $1.8 trillion in cash was posted as collateral against loans and other transactions in 2008, with $1.3 trillion coming in the form of securities and other instruments, according to data provider IHS Markit. A decade later, those positions have inverted, with non-cash collateral balances standing at $1.6 trillion, compared with $870 billion for cash."[0]
and
“If an institution wishes to use [Treasury] assets for financing, to gain yield through lending them out or to meet their HQLA requirements, putting the securities into triparty is the most efficient way to achieve those goals”[0]
Since treasuries (and bonds in general through collateral transformation, emphisis on "non-cash collateral" above meaning not just treasuries) are being used to finance more risk asset purchases.
[0] https://www.bnymellon.com/us/en/what-we-do/markets/aerial-vi...
I don't know about that:
> The Nasdaq 100 ETF (QQQ) is up an astonishing 25.5% this year during a pandemic and that’s including a 29% peak-to-trough drawdown. But the long-term treasury ETF (TLT) is up 27.3%.
* https://awealthofcommonsense.com/2020/08/why-would-anyone-ow...
Though the article goes on to say the primary reason for owns bonds are:
* Bonds hedge stock market volatility.
* Bonds can be used to rebalance.
* Bonds can be used for spending purposes.
* Bonds protect against deflation.
See also his 2018 Q4 article (when the S&P 500 was in the middle of its 20% correction):
* https://awealthofcommonsense.com/2018/10/the-case-for-bonds/
If this is true, it has large and significant ramifications on optimal portfolio construction.