There appear to be two discrepancies in the way interest is computed and credited to TIAA's fixed income "Traditional" accounts. The interest paid is generous by today's standards (2014) however TIAA does not seem to be following industry standards in reporting and computing the interest. The first discrepancy is on the order of a tenth of a percent per year, which does not sound like much, but can amount to thousands of dollars over many years for a large account. The second is not a significant dollar value but makes one wonder if TIAA is using proper computational methods and whether TIAA has adequate controls over its systems.
Specifics:
Issue [1] - Under the heading "view interest rates" on the members account page on its website TIAA shows an interest "crediting rate" for the account. It turns out that this "crediting rate" is not the rate that results in the interest credited on any particular day. It is the APY (annual percentage yield) that results from the interest actually used in the daily calculation. Banks and S&Ls typically state both the interest rate used in their calculations as well as the resulting APY; Presumably this is required by applicable regulations. But banks never advertise the APY as an interest rate. On the TIAA web site one finds a link called "About TIAA Rates" but this link does not take one to anything that would suggest that the advertised rate is not the real interest, that it is actually the APY.
Issue [2] - The daily interest paid varies randomly day to day, over a 40 cent range for a half million dollar account. Over time this more-or-less averages out, but such random variability should not appear in a modern financial institution's computations. It's like the accounts are updated by hand every evening by an ad hoc staff of clerks. Given this variability, the algorithms used in the computations are suspect and one must ask: where is the accounting and audit oversight that would prevent such errors? Does TIAA have competent programmers developing the accounting computations? If the TIAA-CREF website is any indication, they do not. The website is a joke compared to what you see at Schwab.com or Vanguard.com or even Fidelity.com .
Quoted as interest by TIAA: | 4.426% |
True interest, if quoted interest is an APY: | 4.331% |
Date | Days | Account value | Interest actually paid per day | Correctly computed interest using quoted interest - | |
---|---|---|---|---|---|
- as interest | - as APY | ||||
17-Oct-14 | - | 514,027.45 | - | - | - |
20-Oct-14 | 3 | 514,210.49 | 61.01 | 62.35 | 61.02 |
21-Oct-14 | 1 | 514,271.59 | 61.10 | 62.36 | 61.02 |
22-Oct-14 | 1 | 514,332.63 | 61.04 | 62.37 | 61.03 |
23-Oct-14 | 1 | 514,393.45 | 60.82 | 62.38 | 61.04 |
24-Oct-14 | 1 | 514,454.61 | 61.16 | 62.38 | 61.05 |
27-Oct-14 | 3 | 514,637.63 | 61.01 | 62.41 | 61.07 |
28-Oct-14 | 1 | 514,698.83 | 61.20 | 62.41 | 61.07 |
29-Oct-14 | 1 | 514,759.85 | 61.02 | 62.42 | 61.08 |
30-Oct-14 | 1 | 514,821.08 | 61.23 | 62.43 | 61.09 |
31-Oct-14 | 1 | 514,881.92 | 60.84 | 62.43 | 61.10 |
03-Nov-14 | 3 | 515,065.35 | 61.14 | 62.46 | 61.12 |
04-Nov-14 | 1 | 515,126.54 | 61.19 | 62.46 | 61.13 |
05-Nov-14 | 1 | 515,187.42 | 60.88 | 62.47 | 61.13 |
06-Nov-14 | 1 | 515,248.59 | 61.17 | 62.48 | 61.14 |
07-Nov-14 | 1 | 515,309.74 | 61.15 | 62.49 | 61.15 |
10-Nov-14 | 3 | 515,493.31 | 61.19 | 62.51 | 61.17 |
11-Nov-14 | 1 | 515,554.21 | 60.90 | 62.52 | 61.18 |
12-Nov-14 | 1 | 515,615.48 | 61.27 | 62.52 | 61.18 |
13-Nov-14 | 1 | 515,676.65 | 61.17 | 62.53 | 61.19 |
14-Nov-14 | 1 | 515,737.87 | 61.22 | 62.54 | 61.20 |
17-Nov-14 | 3 | 515,921.42 | 61.18 | 62.56 | 61.22 |
18-Nov-14 | 1 | 515,982.64 | 61.22 | 62.57 | 61.23 |
19-Nov-14 | 1 | 516,043.88 | 61.24 | 62.58 | 61.23 |
20-Nov-14 | 1 | 516,105.17 | 61.29 | 62.58 | 61.24 |
21-Nov-14 | 1 | 516,166.34 | 61.17 | 62.59 | 61.25 |
24-Nov-14 | 3 | 516,349.84 | 61.17 | 62.61 | 61.27 |
25-Nov-14 | 1 | 516,411.10 | 61.26 | 62.62 | 61.28 |
26-Nov-14 | 1 | 516,472.36 | 61.26 | 62.63 | 61.28 |
28-Nov-14 | 2 | 516,594.91 | 61.27 | 62.64 | 61.30 |
01-Dec-14 | 3 | 516,778.94 | 61.34 | 62.66 | 61.32 |
02-Dec-14 | 1 | 516,840.21 | 61.27 | 62.67 | 61.33 |
03-Dec-14 | 1 | 516,901.44 | 61.23 | 62.68 | 61.34 |
04-Dec-14 | 1 | 516,962.84 | 61.40 | 62.69 | 61.34 |
05-Dec-14 | 1 | 517,024.01 | 61.17 | 62.69 | 61.35 |
08-Dec-14 | 3 | 517,207.98 | 61.32 | 62.72 | 61.37 |
- | Average | 61.07 | 62.43 | 61.09 |
Issue [1] -
Interest actually paid vs. correctly computed interest:
- on average is short by $1.37 per day
or approaching $500 per year.
And if you look at the entire TIAA Traditional Fund over the years,
TIAA has underpaid its members an estimated 10 to 20 billion dollars interest
as compared to the correctly computed interest for the interest rates advertised.
And that is billion with a b.
Issue [2] -
Variability of daily interest as paid day by day for an actual account,
normalized to half a million dollars at about the start of 2014,
is shown in the following graph:
The "computed" line uses the quoted TIAA interest
as if it were specified as an APY.
Issue [1] - I was sent a pamphlet explaining TIAA rate calculations which does in fact state, but not justify, that the quoted interest rates are not the rates used in the computation but the annual interest that will result from compounding the stated rate for a year. The term APY is not used, but the quoted formula is that generating the APY. This confirms the above statements but does not justify them. I suspect the TIAA Committee Members who approve the rates don't know that their approved rates are treated as APY.
The pamphlet is not present in any of the contractual information related to the TIAA Traditional account. It is not part of the TIAA web site. Another comment I received is that TIAA is not subject to rules that apply to Bank, etc. That TIAA is subject to insurance company rules. This is a legal position that I can't address, but doubt that it justifies presenting the APY as the interest rate.
Issue [2] - It was suggested that the variability is due to the daily NAV computations for the funds backing up the TIAA Traditional account. Although it is quite likely that the NAV does undergo such day-to-day variability, I see no reason why this variability should pass through on a day-to-day basis to individual fixed income accounts. I can't accept this explanation. It is more likely that the computational algorithms are totally obsolete and suffer from truncation errors that carry over from the days before computers.